{"paper_id":"0ed1bdf2-239a-4253-8f32-e671a101b7d3","body_text":"Library transgenesis in zebrafish through delayed site-specific mosaic \nintegration for in vivo pooled screening of transgenes \n \nAuthors: Shahar Bracha1,3,4, Adam Amsterdam5, Yasu Xu1,3,4, Liyam Chitayat1, Anubhav \nSinha1, Edward Boyden12345+ \n \nAffiliations: \n1McGovern Institute for Brain Research, MIT, Cambridge, MA \n2HHMI, Cambridge, MA \n3Yang Tan Collective, K. Lisa Yang and Hock E. Tan Center for Molecular Therapeutics in \nNeuroscience, K. Lisa Yang Center for Bionics, MIT, Cambridge, MA \n4Departments of Brain and Cognitive Sciences, Media Arts and Sciences, and Biological \nEngineering, MIT, Cambridge, MA \n5Koch Institute For Integrative Cancer Research, MIT, Cambridge, MA \n+Contributed by Edward Boyden, edboyden@mit.edu \n \nClassification: Biological Sciences, Genetics \n \nKeywords: protein engineering, neurobiology, synthetic biology, transgenesis, technology \ndevelopment. \n \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nAbstract \nFunctional screening through systematic deletion, editing or addition of libraries of genes is a \npowerful approach for discovering gene functions and developing improved molecular tools. \nHowever, due to the need for high throughput, such campaigns are typically conducted in vitro, \nleading to many discoveries, especially tools and therapeutics, which fail to translate in vivo. \nTissue context, cellular physiology, and systemic regulation shape both tool performance and \ngene function in ways that simplified culture systems cannot predict. Pooled in vivo screening \nmethods have the potential to enable screening within living animals while preserving the \nphysiological context, but current approaches using viral vectors face three critical limitations: \nmulti-transgene insertions per cell confound genotype-phenotype association, viral packaging \nconstrains transgene size, and cell-type tropism restricts and biases targeting. Here, we \nintroduce a zebrafish library transgenesis method that overcomes these limitations through \ndelayed site-specific mosaic integration. We exploit a temporal delay between library \nmicroinjection with PhiC31 mRNA, and library integration, to allows the library to spread \nepisomally throughout the developing embryo before integration begins. This produces mosaic \nanimals where each cell independently integrates one randomly-selected library member, \nenforced by a single genomic AttP landing site. We demonstrate delivery of multi-kilobase \ntransgenes with high library coverage of 1,378-1,989 unique integrants per animal, and single-\ntransgene-per-cell in ~99% of brain cells. This method provides a platform for direct in vivo \nscreening of large transgene libraries with single-transgene precision, with potential applications \nin both biological discovery and tool development. \n \nMain Text \nIntroduction  \nSystematic screening of genetic libraries, which involves testing many perturbations or \ntransgenes in parallel, can critically accelerate the development of molecular tools and \ntherapeutic interventions. Library screening must balance two competing demands: throughput \n(the number of variants that can be tested simultaneously), and predictive accuracy (how \nfaithfully screening conditions represent the biological context where the gene products will \nultimately function). In vitro screening is typically employed due to the accessibility of high \nthroughput assays, but it often involves sacrificing the predictive accuracy of a screen in critical \nways, as in vitro conditions fail to recapitulate the complex cellular and physiological \nenvironments that govern gene function in vivo (1–6). Evidently, many genetically-encoded tools \ndeveloped in vitro have been later found to have diminished or no activity when applied in vivo. \nFor example, genetically-encoded tools can undergo altered processing and trafficking in vivo, \nincluding aggregation or mislocalization to unintended subcellular compartments and tissues, \nwhich does not manifest during in vitro testing, as has been the case for many voltage indicators \n(7, 8) and soma-targeted biosensors (9, 10). Tools optimized in vitro have been found to exhibit \nunexpected off-target effects, non-specificity and deleterious interactions with endogenous \nprocesses that are present in vivo but absent in vitro. Such has been the case for early calcium \nindicators (11, 12), bioluminescent proteins (13), Cas9 (14) and base editors (15). These side-\neffects can even manifest as toxicity or immunogenicity in vivo, sometimes leading to clearance \nof the tool or death of the expressing cells, issues which have been faced for Cas9 (16), Cre \nrecombinase (17), early calcium indicators (12, 18) and some optogenetic tools (19, 20). These \ncontext-dependent differences mean that many candidates identified through in vitro screening \nfail during subsequent in vivo validation, wasting time and resources despite extensive \noptimization efforts (4, 6).  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nScreening directly in living animals would preserve the physiological context necessary for \naccurate prediction of in vivo performance. However, traditional approaches for in vivo testing \ninvolve expressing and phenotyping one transgene in each animal. This one-by-one testing is \nprohibitively slow, expensive, and labor-intensive, in addition to requiring special ethical \nconsideration when screening libraries containing hundreds of variants (2, 3). Researchers have \nthus been limited to either high-throughput in vitro screening, which offers efficiency at the cost \nof predictive accuracy, or low-throughput in vivo testing that accurately represents physiological \ncontext but is impractical for screening at scale. To address this tradeoff, new techniques have \nbeen emerging to enable direct in vivo pooled screening - multiplexed testing of many genetic \nperturbations or transgenes within single animals. This is achieved using transgenesis and \nmutagenesis methods that deliver pools of genetic modifications to a single animal. These \nmethods create mosaic animals, where different cells in the body harbor different genetic \nmodifications. In this way, many variants can be tested simultaneously in each animal, \npreserving the in vivo physiological context while enabling higher throughput than one-by-one \ntransgenic approaches (21). \nCurrent pooled in vivo screening methods have focused on perturbation screens, using libraries \nof gRNAs or siRNAs delivered via viral vectors in rodents (1, 22–24). These methods have been \napplied to study the effects of endogenous genes within different cell types in vivo, by \nassociating their perturbation (by knockout, knockdown, or mutation) with readouts such as cell \nsurvival and proliferation (measured through enrichment or depletion of gRNA/barcode counts in \nbulk tissue sequencing), gene expression profiling (via single-cell RNA-seq) (1, 22–26), and \nmore recently, imaging-based readouts on fixed tissue (via in situ antibody staining or \nfluorescence in situ hybridization) (27). While these approaches have been powerful for \nunderstanding endogenous gene function in diverse cell types within their native in vivo context, \nthey face three fundamental limitations. First, because transgenesis via viral infection follows a \nPoisson distribution, there is an unavoidable tradeoff between the number of transduced cells \nand the proportion of cells transduced with multiple transgenes (24, 25). Cells harboring multiple \ntransgenes confound interpretation by mixing the effects of different genetic modifications, \nintroducing artifacts that are difficult to correct and reducing the effective throughput of the \nscreen. For example, in low-titer lentiviral delivery of gRNA libraries to the cortex of mouse \nembryos, even sparse targeting of <0.1% of cells resulted in 46% of transduced cells containing \nmultiple perturbations (22). Follow-up studies using optimized AAVs with enzymatic integration \nled to similar proportions of multi-transgene cells at 2% tissue targeting (24). Second, viral \nvectors exhibit cell-type-specific tropism that restricts which cell types and tissues can be \ntargeted with sufficient throughput for screening, and biases which cells are targeted for gene \nphenotyping (28). Viral tropism can also depend on cellular state and other factors that may \nvary unpredictably, introducing further uncontrolled biases into screening results (29–31). Third, \nviral packaging constraints limit the size of transgenes that can be screened. AAV vectors, the \nmost commonly used viruses for in vivo gene delivery, can package ~4.7 kb, preventing their \nuse for screening libraries of large protein-coding transgenes and restricting applications to \nsmall transgenes. As a result, to our knowledge, all pooled in vivo screens performed to date \nhave been of small interfering RNAs or gRNA libraries (1, 5, 22–24, 26, 32–36). \nA different approach for library transgenesis, implemented in C. elegans, is TARDIS \n(Transgenic Arrays Resulting in Diversity of Integrated Sequences) (37). This approach involves \ninjecting pooled libraries into the C. elegans gonads and exploiting delayed transgene \nintegration over multiple generations to generate multiple stable transgenic lines. However, \nTARDIS produces single-transgene animals rather than mosaic animals, and requires raising \nmany animals and characterizing them one-by-one (37). While useful for screening in C. \nelegans, it relies on the nematode’s idiosyncratic assembly of injected transgenes into heritable \nextrachromosomal arrays, and requires multiple generations, preventing its generalization to \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nother model organisms. Lastly, the reliance on Cas9-mediated chromosomal breaks and \nhomology-directed repair leads to low efficiency transgene integration, especially for constructs \nlarger than 1-2 kb (38, 39), requiring the use of selectable markers and processing many \nanimals (e.g. heat-shocking and selecting the progeny of 1200 worms for a library of 12 \npromoters in the TARDIS paper (40)). \nIn this work, we introduce a method for pooled library transgenesis that addresses these \nlimitations, implemented in zebrafish using developmentally delayed site-specific integration. \nWe inject 1-cell embryos containing a genomic AttP landing site with a mixed library of AttB-\ncontaining plasmids alongside PhiC31 integrase mRNA. The temporal delay between mRNA \ninjection and integrase translation and maturation allows embryos to complete many rounds of \ncell division before the integrase become active. During this delay, the injected plasmid library \nspreads passively throughout the developing embryo, distributing to all tissues regardless of cell \ntype (addressing the tropism limitation of viral methods). When integration begins, it occurs \nindependently in many cells, with each cell integrating only one randomly selected plasmid from \nits local episomal pool, enforced by the single genomic AttP landing site (addressing the multi-\ntransgene problem). Unintegrated episomal plasmids are subsequently lost through dilution and \ndegradation, becoming undetectable by 3-4 days post-injection (41–43). Because the plasmids \nare delivered by direct microinjection rather than viral packaging, the method accommodates \nlarge multi-kilobase transgenes (addressing the size limitation). This platform for mosaic \ntransgenesis could accelerate screening campaigns of transgenes and genetic perturbation \nlibraries in vivo, while reducing animal use and associated time and labor for direct in vivo \nscreening. \n \nResults \nConceptual design for a library mosaic transgenesis method based on delayed site-\nspecific integration with PhiC31 \nThe utility of a transgenesis method for in vivo pooled screening depends on three critical \nparameters: (1) efficiency (the fraction of cells expressing a transgene), (2) diversity (the \nnumber of different transgenes expressed per animal and the distribution of their abundance), \nand (3) transgene mutual exclusivity (proportion of cells expressing a single transgene). The \nthird parameter is particularly crucial, as cells expressing multiple transgenes create ambiguity \nabout the mapping between transgenes and their phenotype, which confounds screening \nresults. \nA common transgenesis method in zebrafish involves injection of the 1-cell embryo with a \nplasmid containing transposon arms and an mRNA encoding for Tol2 transposase, leading to \nrandom multi-copy genomic integration (44). In theory, this method can be used to deliver \nmultiple transgenes per animal, but the random multi-copy integration mechanism of Tol2 \nmeans that most cells will express multiple transgenes, violating the transgene mutual-\nexclusivity required to support in vivo pooled screening. A recently developed method \nintroduced construct integration using the site-specific bacteriophage integrase PhiC31, instead \nof the transposase, for single-copy genomic integration (45, 46) (Fig. 1A). This method was \ndeveloped to provide a streamlined way to introduce transgenes into validated safe harbor loci, \nreducing experimental variability imposed by genomic position effects. For this purpose, two \nsafe harbor zebrafish lines were developed by the Mosimann lab, named pIGLET14a and \npIGLET24b, which have a single AttP site on chromosome 14 or 24, respectively (46). \nWe reasoned that delivery of the integrase via injection of the PhiC31 mRNA, which was part of \nthe original Lalonde et al. protocol (46), could provide a mechanism for enforcing a temporal \ndelay between DNA introduction and integration. Furthermore, we hypothesized that this \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\ntemporal delay could lead to mutually-exclusive mosaic transgenesis, if a library of plasmids \n(instead of a single plasmid) was injected (Fig. 1B(i)). The exact kinetics of PhiC31 production, \nmaturation and enzymatic activity are unknown, but we hypothesized that multiple embryonic \ncell divisions could happen before it reached levels sufficient for catalyzing plasmid integration \n(Fig. 1B(ii)). We reasoned that if this is the case, then by the time PhiC31 is active, it could \ninduce many parallel and independent integration events in many different cells (Fig. 1B(iii)), \nresulting in a mosaic animal in which different cells integrated different plasmids from the \ninjected library (Fig. 1B(iv)). Importantly, we reasoned that the presence of only one genomic \nAttP landing site will enforce that each cell integrates only one construct, in a mutually-exclusive \nmanner (37, 47). This is because when the integrase catalyzes recombination between the \nsingle genomic AttP site and a plasmid-borne AttB site, it replaces the genomic AttP with AttL \nand AttR sites that cannot serve as substrates for further integration (Fig. 1A).  \n \n \nFigure 1: Illustration of the mosaic library transgenesis method. \n(A) A single-copy genomic AttP landing site (orange) in the pIGLET (‘phiC31 Integrase \nGenomic Loci Engineered for Transgenesis’) line (46), with an exogenously introduced \nplasmid containing an AttB site (blue) and a transgene cassette (green). Once the PhiC31 \nintegrase enzyme (red) is introduced, it catalyzes recombination between the AttP and AttB, \nleading to single-copy genomic integration of the plasmid. \n(B) Schematic of the overall procedure of delayed site-specific library transgenesis. (i) The 1-\ncell embryo is injected with a mixture of plasmids (the transgene library, drawn as circles with \nblue, magenta and green rectangles) and mRNA encoding for the PhiC31 integrase (red). (ii) \nDuring early development, the library passively spreads in the embryo as episomal plasmids \ntogether with the PhiC31 mRNA/protein as the cells divide. (iii) After an initial stage of \ndevelopment, the PhiC31 becomes active and integrates a single randomly-selected plasmid \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nfrom the library in each cell. (iv) This produces a mosaic animal in which different cells \nexpress different library members, and only one library member in each cell.  \n \nPhiC31-mediated mosaic integration leads to high brain targeting with mutually-exclusive \ntransgene expression \nTo test the transgene mutual-exclusivity of the proposed mosaic transgenesis method, we \ninjected heterozygous pIGLET zebrafish embryos with a simplified two-member library of \ntransgenes. The library consisted of a green and a red fluorescent protein (GFP and mScarlet), \neach with a membrane tag. The transgenes were expressed pan-neuronally using a non-\nrepetitive UAS (4xnrUAS) promoter in a HuC::Gal4 driver line (48). We injected this 50:50 \nmixture, together with PhiC31 mRNA, into 1-cell embryos from a cross of pIGLET14a or \npIGLET24b females (carrying a single genomic AttP landing site) and \nHuC::Gal4;nacre;RH1::DsRed males (providing pan-neuronal Gal4 expression required to \nactivate the 4xnrUAS minimal promoter in the integrated constructs). We reasoned that if \nconstruct integration is truly mutually-exclusive, the vast majority of fluorescent neurons will be \neither green or red, and not both. If integration was not mutually exclusive, we would expect to \nsee many cells which are both green and red.  \n3 and 5 days post-fertilization (dpf), we imaged the larvae to assess the distribution of \nfluorescent neurons. This allowed us to estimate both the total number of fluorescent neurons \n(indicating the number of neurons that integrated any construct) and the ratio of single-\ntransgene vs. multi-transgene neurons. We counted fluorescent neurons from eight mosaic \nanimals total. For six of the animals, we counted 1-5 representative FOVs for each (2,473 \nneurons total), and for two animals we counted neurons across the entire hindbrain volume \n(2,511 neurons total) to get an estimate of total integration levels in the brain.  \nOur analysis showed that 99.34% of neurons expressed either GFP or mScarlet exclusively, \nwhile only 0.66% (33/4,984 neurons) expressed both fluorophores (Fig. 2, Table S1). The 33 \ndouble-positive neurons were distributed amongst the eight fish and different brain areas (Table \nS1). Given that the original library consisted of an equal mix of GFP and mScarlet, we reasoned \nthat the probability of integrating both at least one GFP and at least one mScarlet construct \nmust be similar to or larger than the sum of the probabilities of integrating multiple GFP-only \nconstructs or multiple mScarlet-only constructs (which would also appear as GFP-only or red-\nonly fluorescence). Based on this assumption, we estimate the total ratio of neurons that \nintegrated multiple transgenes to be up to double the ratio of red-and-green neurons, translating \nto up to around 1.3% multi-transgene neurons. We hypothesize the rare double-positive cells to \nbe due to spontaneous genomic integration of naked plasmid DNA, a phenomenon known to \noccur at low frequency without enzymatic mediation, with consistent rates reported in the \nliterature (41). This suggests that PhiC31 mosaic integration is indeed mutually exclusive, and \nthe frequency of multi-transgene cells achieved with this method is sufficiently low for high \nquality in vivo pooled screening.  \nWe estimated the total frequency of targeted neurons across the brain based on the counts of \nfluorescent neurons across the entire hindbrain volume of two representative 5 dpf larvae, which \namounted to an average of 1,255 fluorescent neurons per hindbrain (980 in Fish 7 and 1,531 in \nFish 8) (Table S1). The hindbrain was chosen as it enabled the most accurate identification of \nfluorescent neuronal somata, thanks to its structure and distance from the eyes. In more detail: \nthe HuC::Gal4 driver line used contains a multi-copy red fluorescent eye marker (RH1::DsRed) \nas a genotyping aid, which exhibits variable expression levels. Therefore, in some fish, brain \nareas close to the eyes (the optic tectum and forebrain) exhibited a red fluorescent glow, under \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nconfocal microscopy, which interfered with the identification of mScarlet-expressing neurons in \nthose areas. In addition, the high density of axons in the optic tectum around the eyes interfered \nwith the identification of neuronal somata (both green and red) in those areas. We used the \npublished Z brain atlas (zebrafishexplorer.zib.de, (49)) and dataset from Ahrens at al (50, 51) to \nestimate the total number of neurons in the hindbrain of 5 dpf larvae to be around 25k. Given \nthis estimate, our quantification of around 1,255 fluorescent neurons across the hindbrain \nsuggests that approximately 5% of neurons expressed library integrants, assuming that the rate \nof integration was similar in all areas of the brain. The latter was consistent with our qualitative \nassessment of 25 imaged fish over 6 independent experiments (Fig. 2). Those fish showed \nqualitatively similar levels and distribution of fluorescent neurons across the brain.  \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nFigure 2: Mutually exclusive expression of library transgenes. \npIGLET heterozygous larvae display ~99% mutually exclusive mosaic expression of a single \nlibrary member per neuron. (A) Illustration of the experiment: pIGLET heterozygous embryos \ncontaining a single AttP site in chromosome 14 or 24 were injected with PhiC31 mRNA and a \n50:50 mixture of plasmids containing an AttB site (blue) and constructs for neuronal \nexpression of mScarlet (magenta) or GFP (green), fused to a CAAX tag for membrane \ntargeting. 3 or 5 days later, the larvae were imaged. (B-E) Representative images of a mosaic \n5 dpf pIGLET24b heterozygous animal (pIGLET24b;HuC::Gal4;nacre;RH1::DsRed) following \nlibrary transgenesis, showing the forebrain and midbrain (B), midbrain and hindbrain (C), \nposterior hindbrain and spinal cord (D) and spinal cord (E). Max-projection images are shown \nwith skin autofluorescence removed to aid visualization. (F) One plane of the fluorescent \nchannels overlaid on a brightfield image of a mosaic 5 dpf pIGLET14a heterozygous larva \n(pIGLET14a;HuC::Gal4;nacre;RH1::DsRed, with prominent expression of the red eye marker). \n(G) Zoomed-in image of the section marked with a cyan box in the hindbrain in (C), showing a \nneuron co-expressing GFP and mScarlet. (H) Quantification of the ratio of neurons expressing \nboth GFP and mScarlet (double-positives), out of all transduced neurons, in 8 mosaic larvae. \nThe animals for which the whole hindbrain was quantified are marked in orange. The 3 dpf \nlarva is marked in blue. The rest (marked in gray) are 5 dpf larvae for which 1-4 random FOVs \nwere quantified. All the raw numbers are available in Table S1. (I-L) As in B-E, but for a 3 dpf \nlarva. Scale bar: 50 μm. \n \nQuantifying the number and distribution of library transgenes integrated in mosaic \nanimals \nAfter establishing the overall efficiency and mutual exclusivity of library integration, we set out to \nquantify the maximum library diversity achievable per animal. We reasoned that the total \nnumber of different transgenes that can be expressed in one animal will be determined by the \nnumber of independent integration events, which is affected by the developmental stage at \nwhich PhiC31 becomes active. Prior studies following embryos after injection of mRNA \nencoding for GFP showed that significant green fluorescence is detectable at 3 hours post-\nfertilization (hpf) (52). Consistent with that, evidence shows that after injection of PhiC31 mRNA \nto the 1-cell embryo, construct recombination can be observed as soon as 3.3 hpf (45). By 3.3 \nhpf, a zebrafish embryo is estimated to contain 1-2k cells (53). Of course, given the likely \npossibility that integration events are distributed over time, the actual number of independent \nintegration events could be lower if most of them occur before 3.3 hpf, or higher if they continue \noccurring at later stages as well. It also depends on the dynamics of retainment of the PhiC31 \nmRNA/protein and of the episomal library of AttB-containing plasmids as cells divide. The latter \nlikely depends on multiple factors, including: (i) how many plasmid and mRNA molecules were \noriginally injected, (ii) the rate of degradation of the episomal plasmids, mRNA and PhiC31 \nprotein, and (iii) the uniformity of the distribution of the episomal plasmids and integrase \nmRNA/protein among the embryo cells as they divide. \nTherefore, we set out to empirically quantify the number of independent integration events that \noccur in our protocol. We generated a library of plasmids containing 15-nucleotide random DNA \nbarcodes preceding a GFP-CAAX expression cassette (Fig. 3A). This design allowed us to use \nbulk deep sequencing to quantify the diversity of integrated transgenes based on the DNA \nbarcodes recovered from each mosaic animal. We injected the barcoded library into 1-cell \nheterozygous pIGLET embryos (pIGLET24b;HuC::Gal4;nacre;RH1::DsRed) and imaged larvae \nwith widespread neuronal GFP expression at dpf 5 (Fig. 3B). After imaging, we extracted \ngenomic DNA from their entire body and amplified the pool of integrated barcodes using 10-\ncycle PCR with primers spanning the genomic integration junction. The primers were designed \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nto ensure that only barcodes from transgene constructs that integrated into the AttP landing site \ngot amplified, and to exclude the possibility of amplifying barcodes from unintegrated episomal \nplasmids. This was achieved using a forward primer targeting a genomic sequence upstream of \nthe AttP on chromosome 24, and a reverse primer targeting the plasmid-specific HS4 insulator \nsequence downstream of the barcode location. This resulted in a 650 bp amplicon library \ncontaining the integrated barcodes from each larva (Fig. 3C). The amplicon libraries from 12 \nimaged larvae that displayed high levels of neuronal GFP expression and strong amplicon \nbands (Fig. S2) were further processed to generate sequencing libraries. The 650 bp amplicon \nlibrary from these larvae were re-amplified with a 15-cycle PCR, using primers that were internal \nto the first ones, and added illumina-compatible sequencing overhangs and sample-specific 5-nt \nbarcode to the amplicons from each larva, to allow for multiplexed pooled sequencing. Deep \nsequencing yielded a total of 7.2 million high-quality reads across the 12 fish samples. In \naddition, we sequenced the source plasmid library by amplifying the barcodes from the original \ninjected plasmid sample, which yielded 5.8 million sequencing reads (Table S2). The 15-nt \nbarcode was extracted from each read based on a sequence search for the conserved \nsequences surrounding the barcode. We collapsed all closely related barcodes (defined as \nLevenshtein distance of 1 apart), reasoning that given the high complexity of the original library, \nthe likelihood of such closely-related barcodes appearing in sequenced fish by chance is far \nexceeded by the likelihood of 1-nt sequence divergence resulting from sequencing errors or \nmutations introduced during amplification (Table S3, Fig. S4). The fact that most clusters of \nclosely-related barcodes consisted of one high-count barcode and multiple low-count barcodes \nfurther support this assumption. We filtered out rare barcodes that appeared less than 3 times in \na given fish, to further exclude rare barcodes that could represent potential sequencing errors, \ncontaminants, or errors in the barcode extraction. While the proportion of reads with rare \nbarcodes was low in the fish samples (<1% of reads, Table S3), in the source barcode library, \nmost of the reads contained rare barcodes (appearing <3 times in the sequenced library), \nsuggesting that the complexity of the injected library was substantially higher than the \nsequencing depth. In accordance with this observation, we used a more permissive threshold \nfor rare barcode inclusion in the source library (removing only barcodes with count<2, which \naccounted for 45% of the reads). Overall, our analysis revealed that each mosaic animal \nintegrated on average 1,676 different barcodes (median: 1,682, Std. Dev.: 176, range: 1,378-\n1,989) (Fig. 3D, Table S3). As mentioned above, analyzing the original injected barcode library \nrevealed that it had very high complexity, with no barcode being represented at >0.0005% \nfrequency in the library even after collapsing closely-related barcodes. This makes it likely that \nfor each fish, each unique barcode observed originated from a single integration event.  \nBarcode sequence analysis revealed no favored sequence composition or motifs for the \nbarcodes from the integrated plasmids compared to the injected library (Fig. 3F, Fig. S3). \nFurthermore, the sequence diversity, measured by the average pairwise hamming distance \nbetween the barcodes integrated in each fish was the same as that of the injected library (11.1), \nand similar to the expected value for a theoretical uniform random sequence library (11.3), \nindicating that the different barcodes identified in each fish are indeed random and are the \nproducts of independent integration events, rather than from barcode diversification by mutation \nof a small number of integrants in each fish (Fig. S4). \nWhen looking at the distribution of barcode representation, we found clear signs of clonal \nexpansion of the barcodes within the mosaic animals, as expected. While the sequenced source \nlibrary displayed millions of different barcodes, each represented at similar rare frequency, the \nmosaic animals tended to have a different set of around 1,600 barcodes each, and the barcodes \nwere represented at different frequencies (Fig. 3E, Fig. S5, Fig. S6). As the high library \ncomplexity suggests that each unique barcode originated from a single integration event, \nvariance in barcode abundance is most likely the result of clonal amplification as cells replicated \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nafter barcode integration, rather than multiple independent integration events of the same \nbarcode. After a cell integrates a library member, all its progenies will inherit the same barcode, \ncreating clonal populations whose size reflects the number of cell divisions between integration \nand gDNA extraction at 5 dpf. Therefore, the variance in integrated barcode abundance could \nrepresent either (1) variable integration timing - barcodes integrating earlier in development \nundergo more amplification with cell division; or (2) variable proliferative capacity of different cell \nlineages- namely, barcodes integrating into highly proliferative lineages (e.g. basal stem cells) \nachieving greater expansion than those in slowly dividing or post-mitotic cell types (e.g. \nneurons). Accordingly, we would expect that the variance in transgene representation would be \nlower in library transgenesis applications involving cell type- or tissue-specific expression (for \nexample, when screening transgenes only in neurons). In such applications, the total number of \ndifferent transgenes expressed per animal would also be smaller, since our DNA barcode \nquantification included all barcodes integrated across the entire body of the zebrafish.  \n \n \nFigure 3: quantifying the capacity of delayed integrase library transgenesis with DNA \nbarcodes \n(A) Construct design for the barcoded GFP-CAAX library plasmids. Each plasmid includes an \nAttB sequence (light blue), HS4 insulator element (medium gray) embedded with a unique 15-\nnt random sequence barcode (15xN, purple), a 4xnrUAS minimal promoter for tissue-specific \nGal4 transcription activation (light gray), GFP-CAAX (green arrow) and SV40 polyA (dark \ngray). (B) Scheme illustrating the experiment- a high-complexity library of barcoded GFP-\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nCAAX plasmids, each containing a different 15xN barcode (dark blue, magenta and yellow \nrectangles on the plasmids), was injected into 1-cell embryos of heterozygous pIGLET24 \nzebrafish containing a genomic AttP site on chromosome 24. The library was co-injected \ntogether with mRNA encoding for the PhiC31 integrase (red). 5 days later, the larvae were \nimaged to confirm GFP expression in neurons, and then selected for extraction of genomic \nDNA from their entire bodies (n=12 larvae). (C) the genomic extracts were used as templates \nfor 10-cycle PCR amplification of the barcodes from all genomically integrated plasmids, using \nprimers that specifically target the integration junction (shown in red arrows). The resulting \n650-bp amplicon library is then re-amplified with different, internal primers (shown in blue \narrows), to attach overhangs for Illumina next generation sequencing (blue) and add a 5-nt \nsample-specific multiplexing barcode (pink) for pooled sequencing. (D) Number of unique \nhigh-confidence barcodes identified in each fish, after barcode collapsing and filtering. (E) \nHistogram of barcode abundance for the injected source library and for the barcodes \nrecovered from the fish-integrated plasmids. The injected library displays a narrow \ndistribution, indicating high complexity (many rare barcodes appearing at similar low \nfrequency) while the fish-recovered barcodes display a broader long-tailed distribution (some \nbarcodes appearing much more than others), consistent with intra-fish clonal expansion of the \nintegrated transgenes. The per-fish barcode abundance histograms are available in Fig. S5. \nRPM=reads per millions (read counts normalized to the total number of reads sequenced for \neach sample). (F) Nucleotide composition for each position in the injected library and in the \nintegrated barcodes shows high sequence diversity and no sequence bias for integration. Per-\nfish barcode sequence compositions are available in Fig. S2.  \n \nDiscussion \nWe developed a method for library transgenesis that provides a platform for high-throughput in \nvivo screening. By exploiting a temporal delay between plasmid library injection and PhiC31-\nmediated integration in zebrafish, we achieve mosaic transgenesis with 1,378-1,989 unique \nintegrated transgenes per animal and ~99% mutually-exclusive transgene expression. Site-\nspecific integration to a single genomic landing site ensures that nearly all transduced cells \nexpress a single transgene while still enabling delivery to many cells across the tissue, \ncircumventing the tradeoffs that limit methods relying on stochastic infection and integration \nevents such as viral delivery, transfection and random transposase-mediated genomic \nintegration. By creating mosaic animals in which different cells express different transgenes, we \ncan effectively transform each animal into hundreds of parallel experiments. This could enable \nhigh-throughput screening of transgene libraries in the native in vivo physiological context, \nwhich would be valuable for developing better genetically-encoded tools through direct in vivo \nscreening, and for basic research investigating the effects of libraries of genetic perturbations in \nvivo. \nWe demonstrate mosaic somatic transgenesis, where libraries of transgenes are expressed in \ndifferent somatic cells following delayed integration, allowing each animal to function as a living \nlibrary with individual cells testing different variants in the native in vivo context. This approach \nis well-suited for screening genetic perturbations and transgenes with cell-autonomous effects, \nwhere the phenotype of a single transduced cell can be reliably assessed even when \nsurrounded by non-transduced or differentially-transduced neighbors. Applications include \nscreening molecular tools in vivo, such as genetically-encoded biosensors, fluorescent markers, \nDNA editing enzymes and more. Additionally, mosaic somatic transgenesis could be applied to \nenable lineage tracing and brainbow-like barcoding strategies for morphological tracing and cell \nsegmentation, including for connectomic mapping (54–57). Beyond this demonstrated \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\napplication, we hypothesize that mosaic germline transgenesis could also be achievable, where \ninjected animals would generate libraries of progeny animals, each containing a single \ntransgene throughout its entire body, similar to the TARDIS approach (37).  \nWhile we demonstrate mosaic integration of 1,378-1,989 unique library variants per animal, \nfuture optimization could further increase library complexity and integration efficiency. Protocol \nrefinements may include varying the concentrations and purification methods used for the \ninjected plasmid libraries and mRNA and refining the AttB plasmid design. Using optimized \nhyperactive PhiC31 instead of the native bacteriophage sequence could provide another \nstrategy to enhance integration efficiency (59). It would be interesting to investigate more deeply \nthe timing of integrase expression and activity, and to explore alternative mechanisms for \ndelayed integration. One option could involve introducing the integrase gene as DNA instead of \nmRNA. For example, if the integrase was encoded on a co-injected DNA plasmid under a \nubiquitous promoter, its transcription would only begin around 3 hours post-fertilization with the \nstart of zygotic transcription after 10 embryonic cell divisions (60), likely resulting in an even \nlonger temporal delay before integration. Alternatively, a tissue-specific promoter could further \ndelay integration until after a specific tissue or cell lineage forms, while an inducible promoter \n(e.g., heat-shock or drug-inducible) would provide flexible spatiotemporal control of integration \ninitiation (61–63). Future approaches could involve generating transgenic zebrafish lines with \ndelayed or inducible integrase expression cassettes in their genome, further increasing overall \nefficiency and tissue coverage. \nIn our current implementation, DNA barcode library analysis demonstrated clonal expansion of \nthe integrated variants, which was expected from the mechanism of delayed integration into \ndifferent cells across the entire body. We hypothesize that this could result from either variance \nin integration timing or variance in the proliferative capacity of different cell types receiving \ndifferent integrants. If the latter is the case, we would expect this variance to be significantly \nlower when the method is applied to screening library variants expressed in a specific tissue or \ncell type. In addition, more precise control of the variant abundance distribution could be \nachieved by keeping the injected library smaller than the number of integration events. The \noverall library complexity (measured both by the number of unique variants and their \ndistribution) achieved here should be suitable for many screening applications, and further \nimprovements could enhance the method’s throughput and efficiency. \nWe implemented this approach in zebrafish, which offers unique advantages for many of the \napplications we discuss. Its natural transparency and small size make it highly amenable to \nimaging-based phenotypic analysis of transgene and perturbation libraries in live mosaic \nanimals. As a vertebrate model, it recapitulates many aspects of human physiology with \ndemonstrated clinical translatability and a wealth of established disease models (64, 65). \nCritically, this work was enabled by the existence of AttP landing site lines with validated safe-\nharbor integration sites that were already established for zebrafish (46). Similar engineered lines \nwith genomic integrase landing sites have also been established for C. elegans, Drosophila, \npigs and mice (66–71), providing a foundation for adapting this approach to those model \norganisms as well. As we increasingly recognize that gene functions depend critically on their \ninteractions within complex in vivo environments, including aspects we may not yet even fully \nunderstand, methods that preserve this physiological context while enabling high-throughput \nscreening will be essential for both developing better molecular tools and for basic biological \ndiscovery. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nAcknowledgements \nWe thank Christian Mosimann and his lab for the pIGLET zebrafish lines. Thanks to all \nmembers of the Boyden lab for many fruitful discussions. Fig. 1B, 2A, 3B contain illustrations \nfrom biorender.com. ESB acknowledges, for funding, Lisa Yang, HHMI, NIH 1U01NS120820, \nNIH 1R01MH123977, NIH R01MH122971, and NIH R01DA029639. SB acknowledges funding \nfrom the Y. Eva Tan Postdoctoral Fellowship and the Yang Tan Collective at MIT.  \n \nMethods \nMaterials and data availability \nAll the raw imaging data and Illumina sequencing data associated with Fig. 2, 3, S1, S2, S3, S4, \nS5, S6, and Table S1, S2, S3, are available on DOI https://doi.org/10.5061/dryad.d2547d8h0. \nAll the code used to analyze the sequencing data is available on \nhttps://github.com/shaharbr/library_transgenesis. The full sequences of key plasmids and \nprimers used in this study are available in appendix data S1, S2, S3 and S4. The plasmids AttB-\nHS4-nrUAS-GFP-CAAX and AttB-HS4-nrUAS-mScarlet-CAAX will be made available through \nAddgene upon publication. \nZebrafish husbandry and transgenesis \nAll procedures were done in accordance with government and university guidelines, and \napproved by the MIT Committee on Animal Care. Heterozygous pIGLET embryos for injection \nwere obtained by crossing homozygous pIGLET14a or pIGLET24b or \npIGLET24b;HuC::Gal4;RH1::DsRed;nacre (for the single-copy AttP landing site) with \nHuC::Gal4;nacre (for HuC-driven pan-neuronal expression of proteins under the minimal \n4xnrUAS promoter) adult zebrafish. Homozygous pIGLET14a and pIGLET24b were obtained as \na gift from Prof. Chris Moismann’s lab (46). HuC::Gal4;RH1::DsRed;nacre;nacre zebrafish were \ngenerated in-house based on HuC::Gal4;RH1::DsRed obtained as a gift from Prof. Herwig \nBaier’s lab. Homozygous pIGLET14a;HuC::Gal4;RH1::DsRed;nacre and \npIGLET24b;HuC::Gal4;RH1::DsRed;nacre lines were generated in-house by crossing the \nabove. 1-cell embryos were microinjected with approximately 1 nanoliter of injection mix \ncontaining 50 ng/μL total plasmid DNA and 50 ng/μL PhiC31 mRNA, mixed in a total volume of \n5 μl RNAse-free water with 0.1% phenol red as a visual marker for successful injection. \nMicroinjections were performed using pulled glass capillaries. Embryos were raised at 28°C in \naquarium makeup water (Instant Ocean solution diluted to 450 microSiemens and adjusted to \npH 7.0 with sodium bicarbonate) until their analysis at 3-7 days-post-ferlitization (dpf). \nPlasmid and mRNA purification \nAll plasmids for microinjection were extracted using the QIAprep Spin Miniprep Kit (Qiagen cat \n#27106) without the addition of RNAse in the lysis buffer, and eluted in water. The PhiC31 \nplasmid (Addgene #68310) was used to produce purified PhiC31 mRNA using the mMESSAGE \nmMACHINE™ T7 Transcription Kit (Thermo Fisher, AM1344) with lithium chloride purification. \nStocks were diluted to 100 ng/μl, aliquoted to 3 μl per tube and kept in the -80C freezer until the \nexperiment. mRNA aliquots were thawed and kept on ice for each experiment, with up to one \nfreeze-thaw cycle per aliquot. \nGeneration of the barcoded plasmid library \nTo generate the DNA barcode library (barcoded AttB-HS4-15N-nrUAS-GFP-CAAX), we added \n15-nt random sequence barcodes into a base plasmid encoding for expression of membrane-\ntargeted GFP (AttB-HS4-nrUAS-GFP-CAAX). The base plasmid contained an AttB site followed \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nby a HS4 insulator sequence, 4xnrUAS (non-repetitive UAS) minimal promoter (48), GFP fused \nto CAAX membrane targeting motif and a polyA sequence, in a pTwist backbone.The GFP \nexpression was used as validation to confirm successful injection and integration in the larvae \nthat were picked for genome extraction and sequencing. The 15-nt random sequence DNA \nbarcodes were added in the middle of the HS4 insulator sequence using 2-cycle PCR with a \nreverse primer containing 15 random bases (N’s), followed by a NheI restriction site and extra \nhomology handles for further amplification (primers F_HS4_NheI and R_Add15N_HS4_NheI in \nAppendix data S4). The 2-cycle PCR was performed on 20 fmol plasmid (~50 ng) with 400 fmol \nof each of the primers (20-fold molar excess), with Q5 high fidelity polymerase mix (cat #M0492 \nNEB). The reactions were incubated in a thermocycler with initial denaturation (98C, 30 sec) \nfollowed by 2 cycles of: 98C for 10 sec, 69C for 30 sec, 72C for 120 sec, and then final \nextension (72C, 2 min). The products were incubated with 1 μl of the restriction enzyme DpnI for \n2 hours at 37C (to eliminate the template plasmid), followed by heat inactivation for 20 min at \n80C, and column extraction. The purified products were then re-amplified with primers \nhomologous to the ends of the 2-cycle PCR primers (primers F_amp_HS4 and R_amp_HS4 in \nAppendix data S4). PCR was performed with Q5 high fidelity polymerase mix, with initial \ndenaturation (98C, 30 sec) followed by 30 cycles of: 98C for 10 sec, 67C for 30 sec, 72C for \n120 sec, and then final extension (72C, 2 min). The products were then incubated for 2 hours at \n37C with 1 μl each of the restriction enzymes DpnI and NheI (to expose the sticky-ends at the \nends of the amplicons), followed by heat inactivation for 20 min at 80C. Then, the products were \ngel extracted and eluted in 20 μl water. Half (10 μl) of the eluted product was then circularized \nby ligation with T4 ligase (cat #M0202 NEB) for 15 min in room temperature, followed by heat \ninactivation at 65C for 10 minutes. 5 μl from the resulting ligation product was transformed into \ne. coli as a pooled library. 1% of the transformed bacteria were plated onto an agar plate with \n100µg/mL carbenicillin, and the remaining 99% was grown in a 40 ml liquid culture of LB with \n100µg/mL carbenicillin overnight. The liquid cultures were used to midi-prep the library using the \nQIAGEN Plasmid Plus Midi Kit (Qiagen cat #12943) without the addition of RNAse in the lysis \nbuffer, and eluted in water. Successful cloning and library complexity quality controls were \nestimated by individual whole-plasmid sequencing of 5 random colonies from the plated \ntransformed bacteria, and by nanopore sequencing of 10,000 plasmid reads.  \nImaging and analysis of the mosaic larvae expressing fluorescent protein libraries \nThe images shown in Fig. 2 and Fig. S2 were acquired of 3-5 dpf mosaic larvae, which were \ninjected with a 50:50 mixture of plasmids for pan-neuronal expression of GFP-CAAX or \nmScalet-CAAX (AttB-HS4-nrUAS-GFP-CAAX and AttB-HS4-nrUAS-mScarlet-CAAX) as 1-cell \nembryos.  \nThe images shown in Fig. S1 were acquired of 5 dpf mosaic larvae injected with the plasmid \nlibrary of barcoded GFP-CAAX (barcoded AttB-HS4-15N-nrUAS-GFP-CAAX) as 1-cell embryos.  \n3 dpf larvae were imaged live, while 5 dpf larvae were fixed in 4% PFA overnight, washed three \ntimes in aquarium makeup water and mounted in 1% low-melting agarose. Mounted fish were \nimaged using a spinning disk confocal microscope (Yokogawa CSU-W1 Confocal Scanner Unit \non a Nikon Eclipse Ti microscope) with 10x air objective and a 40x water immersion objective \n(Nikon MRD77410). The microscope is equipped with a Zyla PLUS 4.2 Megapixel camera \ncontrolled by NIS-Elements AR software, and laser/filter sets for 405 nm, 488 nm, 561 nm and \n640 nm optical channels. We acquired 2-4 FOV for each fish, each as a z-stack with 2.5 μm \nintervals, to cover most of the brain volume. The images shown in Fig. 2B-F and Fig. S1 are \nrepresentative of results obtained over six repeats of the experiment, with 25 larvae total \nimaged aged 3-7 dpf. The images shown in Fig. 2, S1 and S2 are max-intensity projections, \ngenerated using the ImageJ Z Project plugin. For Fig. 2 and S1, green autofluorescence from \nthe skin was removed using manually-drawn masks on the individual z-planes, before \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\ngenerating the max-intensity projection, as demonstrated in Fig. S2. This was done to prevent \nthe obstruction of neurons in one plane by skin autofluorescence in adjacent planes, which \nwould otherwise cover and hide them in the max-projection. All the raw imaging data (.nd2 \nhyperstacks) associated with Fig. 2, S1, S2 and Table S1, are available on are available on \nhttps://doi.org/10.5061/dryad.d2547d8h0. The cell counting quantification in Table S1 was \nperformed on eight larvae aged 3 or 5 dpf, from 6 different injection clutches over 3 independent \nexperiments. Counting of red and green fluorescent neurons was performed manually using the \nImageJ CellCounter plugin.  \nDNA extraction from zebrafish \nFor the results shown in Fig. 3D-F, Fig. S1, S3, S4, S5, S6 and Tables S2 and S3, zebrafish \nlarvae were evaluated for GFP expression at 5 dpf, and after confirmation of GFP expression \n(Fig. S1), 24 were picked for euthanization and lysis for genomic extraction. The larvae were \nlysed in 180 μl Qiagen buffer ATL (cat #19076) with 20 μl proteinase K (20 mg/mL, cat #19134) \nand incubated in 56C for 1 hour with vortexing every 20 minutes, followed by 90C for 20 \nminutes. The lysed samples were then processed with the QIAamp DNA FFPE Tissue Kit (cat \n#56404) using the manufacturer's protocol, starting from the lysis section. The final genomic \nDNA was eluted in 30 μl water.  \nAmplicon library generation from zebrafish amplified integrated barcodes and from the \nsource injected library \nThe genomic extracts from 24 larvae were amplified to produce a 652 bp amplicon of the \nintegrated plasmids (illustrated in Fig. 3C). These amplicons were generated by PCR with a \nforward primer on a genomic location on chromosome 24, ~340 bp upstream of the AttP landing \nsite (F_Chr24pIGLET) and a reverse primer on the plasmid in the HS4, ~200 bp after the \nbarcode (R_HS4), for 10 cycles, with Q5 high fidelity polymerase, using the following conditions: \ninitial annealing with 98C for 30 s, 10 cycles of: 98C for 10 s, 70C for 30 s, 72C for 40 s, and \nfinal extension with 72C for 2 min. We used 10 μl (a third) of each genomic extract as template. \nThen, the PCR products were run on a gel and the 652 bp amplicons were extracted from the \ngel and eluted in 25 μl water. At this point, we selected 12 of the 24 fish-derived amplicon \nsamples for further processing, based on the density of their amplicon bands on the gel, and \nbased on the high level of neuronal GFP expression recorded in the larvae they originated from \n(Fig. S1). 20 μl of each purified amplicon library was used as template for a second PCR \nreaction using primers internal to the first, which also added 5-nt sample multiplexing barcodes \nand Illumina sequencing overhangs in the forward and reverse direction (F_Chr24_illumread \nand R_HS4_FishX_illumread). This PCR went for 15 cycles using similar conditions to the \nabove, and after it the 325 bp amplicons were run on a gel and purified as above. The resulting \npurified amplicons from the different barcoded samples were then combined into one pooled \nsequencing library (illustrated in Fig. 3C). To generate sequencing reads from the source \ninjected plasmid library, we amplified the original plasmid sample with primers that bind \nupstream and downstream of the 15xN barcode in the HS4 element, while adding Illumina \nsequencing overhangs (F_HS4_illumread and R_HS4_illumread). This produced 336 bp \namplicons for direct Illumina sequencing, with similar PCR conditions and purification as the \nabove. \nLibrary prep and Illumina sequencing  \nSequencing libraries were constructed from amplicon samples using the Illumina DNA Prep \ntagmentation kit paired with Illumina Unique Dual Indexes, without the tagmentation steps. \nLibraries were sequenced by SeqCoast Genomics (Portsmouth, NH) on Illumina NextSeq2000 \nusing a 300-cycle XLEAP-SBS flow cell kit, generating paired-end reads (2x150). To ensure \naccurate base calling, 1-2% PhiX control DNA was added to each sequencing run. Post-\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nsequencing processing, including sample demultiplexing, trimming, and run metrics analysis, \nwas conducted using the integrated DRAGEN v4.2.7 software on the NextSeq2000 platform. \nQuality assessment (shown in Table S2) was performed at two levels: evaluation of overall run \nperformance to confirm sequencing data integrity, and targeted review of FastQC quality reports \nfor individual samples. Overall, sequencing of the fish-derived amplicons produced 7.2 million \npaired-end reads total and the source library amplicons produced 5.8 million paired-end reads \ntotal, with >96% bases with Phred quality score >= 30. After demultiplexing, each fish-derived \nsample had 538-643k reads (Table S3). \n \nDemultiplexing and barcode extraction \nRaw sequencing reads were processed using a custom Python-based analysis pipeline, \navailable at https://github.com/shaharbr/library_transgenesis. For the 12-fish pooled library, \nreads were first demultiplexed based on the 5-nt sample barcodes (with tolerance for 1 \nmismatch) and reverse-complemented to correct for sequencing orientation. The injected \nbarcode library reads required no demultiplexing and were processed in their original \norientation. For both samples, the 15-nt variable barcodes were extracted by identifying \nconserved anchor sequences flanking the barcode region: a 12-nt sequence \n(AGCCCCCAGGGA, allowing 2 mismatches) upstream and a 5-nt sequence (CACGC, requiring \nexact match) downstream. The extraction algorithm employed progressive search stringency \n(exact match → Hamming distance up to 2 → Levenshtein distance up to 2) with position \nvalidation to ensure accurate barcode identification. The full results from this analysis are \nincluded in Table S3. \n \nBarcode collapsing, error correction and high-confidence barcode filtering \nTo account for PCR and sequencing errors, barcodes differing by a Levenshtein distance of 1 \n(single nucleotide substitution, insertion, or deletion) were collapsed into a single parent \nbarcode. The parent barcode was defined as the most abundant sequence. All read counts from \nchild barcodes were aggregated into their respective parent barcodes, preserving per-sample \ninformation. \nFor the injected barcode library, barcodes were retained if they had ≥2 reads. For the integrated \nbarcodes from fish, filtering was performed on a per-fish basis: a barcode was retained in a \ngiven fish only if it had ≥3 reads in that fish; otherwise, that fish's count for that barcode was set \nto zero. Barcodes with Levenshtein distance ≤2 to any conserved (non-barcode) region of the \ntemplate read structure were removed to eliminate potential artifacts from faulty barcode \nextraction. The full results from this analysis are included in Table S3. \n \nBarcode abundance and diversity analysis \nRead counts for each barcode were normalized to Reads Per Million (RPM) to account for \ndifferences in sequencing depth between samples (histograms shown in Fig. 3E and Fig. S5). \nFor each sample, we calculated the coefficient of variation (CV), Shannon diversity index (using \nlog base 2), and quartile ratio (Q3/Q1) as measures of barcode abundance distribution (shown \nin Fig. S6). Sequence composition bias was assessed by calculating positional nucleotide \nfrequencies across all barcodes and comparing library and integrated barcode populations \n(shown in Fig. 3F and Fig. S3). Pairwise Hamming distances were computed on 20,000 \nrandomly sampled barcode pairs to quantify sequence diversity (shown in Fig. S4).  \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nSupplementary Information (SI files) \n \n \n \n \n \nFigure S1 - GFP-CAAX expression in the 12 fish with DNA barcodes characterized by deep \nsequencing (related to Fig. 3). Images presented are max projections from confocal \nfluorescence imaging of the brain around the optic tectum and/or hindbrain of the animals, \nwith skin autofluorescence removed with manual masks to aid visualization. Magenta \nfluorescence corresponds to red eye marker expression in the HuC::Gal4;nacre;RH1::DsRed \ndriver fish line. Scale bar = 50 μm. \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\n \nFigure S2 - Demonstration of the skin autofluorescence removal from the max-projections \nimages shown in Fig. 2 and S1. Green autofluorescence from the skin was removed using \nmanually-drawn masks on the individual z-planes, before generating the max-intensity \nprojections. This was done to prevent the obstruction of neurons in one plane by skin \nautofluorescence in adjacent planes, which would otherwise cover and hide them in the max-\nprojection. Top (A-D): max-projections of confocal images from forebrain and midbrain (A), \nmidbrain and hindbrain (B), posterior hindbrain and spinal cord (C) and spinal cord (D), as \nshown in Fig. 2, before and after removal of the skin autofluorescence. Bottom (E-F): Two \nexamples of individual z-planes from the max-projection shown in (B), before and after \nremoval of the skin autofluorescence. Scale bar = 50 μm. \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\n \n \nFigure S3 - Nucleotide composition for each position in the set of barcodes recovered \nfrom each fish, related to Fig. 3. DNA sequence logos showing the positional nucleotide \nfrequencies across all unique barcodes recovered from each of the 12 individual fish. Only 15-\nnt barcodes were included in the frequency calculations, although an additional minority of \nbarcodes were 14 or 16 nt long (<1%). For each position, the height of each colored segment \nrepresents the proportion of barcodes containing that nucleotide at that position. The number \nin parentheses in the subtitle for each plot indicates the total number of unique barcodes \nanalyzed for that fish sample. The close to uniform nucleotide frequencies across all positions \nshow a similar lack of sequence bias in the recovered barcode population for all the animals \nanalyzed. \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\n \nFigure S4 - Comparison of the distribution of pairwise Hamming distances for the sets of \nbarcodes integrated in each fish, the original injected library, and a theoretical library of \nuniformly distributed random 15-nt barcodes. Distributions of pairwise distances are shown for \n20,000 random pairs taken from each set. On top of each violin plot there is an overlaid \nboxplot, showing the median and interquantile range (IQR, representing 25th and 75th \npercentiles), and whiskers extending to 1.5×IQR beyond the quartiles. The number above \neach violin plot shows the mean Hamming distance in each sample. The narrow distribution \ncentered around Hamming distance 11.1 (close to the theoretical maximum for random \nsequences) indicates that barcodes are highly dissimilar to each other, confirming minimal \nsequence clustering or bias in the samples. Furthermore, it confirms that the integrated \nbarcodes from the mosaic fish retained the same sequence diversity as the original injected \nlibrary. \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\n \n \nFigure S5 - Distribution of abundance for the set of barcodes recovered from each fish, \nrelated to Figure 3. RPM=reads per million. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\n \nFigure S6 - Abundance distribution metrics for the injected library and integrated barcodes \nrecovered from each fish, quantifying the extent of clonal expansion heterogeneity and the \nproliferative differences among cells that received different barcode integrations. The number \nabove each bar plot shows the value for that sample. Overall, higher CV, lower Shannon \ndiversity and higher quartile ratios show an increase in skewness of the abundance \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\ndistribution of the barcode represented in each fish, consistent with intra-fish clonal expansion \nof integrated barcodes.  \nThe CV (coefficient of variation) is calculated as the standard deviation divided by mean of \nbarcode read counts. A higher CV means some barcodes are much more abundant than \nothers (uneven distribution), while a lower CV indicates more uniform barcode \nrepresentation.Shannon diversity is an entropy-based metric that considers both the number \nof unique barcodes and the evenness of their distribution (in bits). It quantifies the overall \nbarcode diversity, accounting for both richness (how many different barcodes) and evenness \n(how uniformly distributed their abundances are). Higher values indicate more diverse, more \nevenly distributed barcode populations. The quartile ratio is calculated as the ratio of the 75th \npercentile to the 25th percentile of barcode abundance, quantifying the spread of the middle \n50% of barcode abundances. Higher ratios indicate greater inequality in barcode \nrepresentation. \n \n \nTable S1 - Ratio of multi-transgene neurons in the brains of mosaic zebrafish. \nQuantification of neurons expressing GFP-only, mScarlet-only, or both fluorescent transgenes in \nmosaic transgenic zebrafish brains. Z-plane: The specific optical section(s) analyzed from the \nconfocal Z-stack, indicated as the plane number out of the total stack depth (e.g., \"30/64\" \nmeans plane 30 from a 64-plane stack). Brain region(s) visible in each plane are indicated in \nparentheses. For Fish 7 and 8, the entire hindbrain volume was analyzed rather than a single \nplane. GFP/mScarlet only: Number of neurons expressing only the GFP or mScarlet \ntransgene. Both: Number of neurons co-expressing both the GFP and mScarlet transgenes. \nAll: Total number of transgene-positive neurons counted (GFP only + mScarlet only + Both). \nRatio (Both/All): Percentage of all transgene-positive neurons that express both fluorophores, \ncompared to all counted neurons. This ratio is used to estimate the frequency of multi-transgene \nintegration events. Combined: Summary statistics pooling all the analyzed planes across all \nfish. \nFish ID Age pIGLET \nline \nZ-plane GFP \nonly \nmScarlet \nonly \nBoth All Ratio \n(Both/All) \nFish 1 5 dpf 14a 30/64 (hindbrain \nand forebrain) \n413 93 0 506 0.00% \nFish 1 5 dpf 14a 1/19 (spinal \ncord) \n89 51 3 143 2.10% \nFish 2 5 dpf 14a 16/60 (hindbrain) 51 11 1 63 1.59% \nFish 2 5 dpf 14a 33/60 (hindbrain) 95 56 0 151 0.00% \nFish 3 5 dpf 14a 28/62 (hindbrain \nand forebrain) \n189 57 1 247 0.40% \nFish 4 5 dpf 14a 12/79 (hindbrain \nand midbrain) \n43 22 0 65 0.00% \nFish 4 5 dpf 14a 17/79 (hindbrain \nand midbrain) \n56 41 1 98 1.02% \nFish 4 5 dpf 14a 23/79 (hindbrain 80 35 0 115 0.00% \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nand midbrain) \nFish 4 5 dpf 14a 30/79 (hindbrain \nand midbrain) \n48 35 1 84 1.19% \nFish 5 5 dpf 14a 20/63 (hindbrain \nand midbrain) \n30 15 0 45 0.00% \nFish 5 5 dpf 14a 25/63 (hindbrain \nand midbrain) \n64 18 0 82 0.00% \nFish 6 3 dpf 24b 1 to 15/89 \n(forebrain) \n79 36 0 115 0.00% \nFish 6 3 dpf 24b 36/89 (forebrain) 94 40 0 134 0.00% \nFish 6 3 dpf 24b 12/68 (hindbrain) 88 53 1 142 0.70% \nFish 6 3 dpf 24b 21/68 (hindbrain) 155 76 3 234 1.28% \nFish 6 3 dpf 24b 27/68 (hindbrain) 159 88 2 249 0.80% \nFish 7 5 dpf 14a Entire hindbrain \nvolume \n748 222 10 980 1.02% \nFish 8 5 dpf 24b Entire hindbrain \nvolume \n1030 491 10 1531 0.65% \nCombined: 3511 1440 33 4984 0.66% \n \n \nTable S2 - Sequencing library quality metrics. Quality metrics for the Illumina sequencing \nlibraries from the injected barcode library and pooled integrated barcodes extracted from 12 \nindividual fish. Mean quality score: Average Phred quality score across all base calls in the \nlibrary. The Phred quality score is a logarithmic measure of base calling accuracy, calculated as \nQ = -10 × log₁₀(P), where P is the probability of an incorrect base call. Q20: 1% error rate (99% \naccuracy), Q30: 0.1% error rate (99.9% accuracy). Bases ≥ Q20: Percentage of sequenced \nbases with Phred quality score of 20 or higher. Bases ≥ Q30: Percentage of sequenced bases \nwith Phred quality score of 30 or higher. \n Injected \nbarcode library \nIntegrated barcodes \n(12 pooled fish \nsamples) \nSequenced reads 5,822,820 7,191,895 \nMean quality score: 39.13 39.22 \nBases >= Q20: 99.13% 99.02% \nBases >= Q30: 95.24% 95.84% \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nTable 3 - barcode counts and read retention throughout the stages of barcode extraction \nand processing. Reads, after demultiplexing: Total number of paired-end sequencing reads \nassigned to each sample. For fish samples, reads were demultiplexed based on 5bp sample \nbarcodes at the read start, allowing up to 1 mismatch for error correction. The injected library \nwas sequenced separately and required no demultiplexing. Reads, after barcode extraction: \nNumber of reads from which valid barcodes were successfully extracted. Extraction required \nidentifying conserved anchor sequences flanking the random barcode region. Reads lacking \nproper anchor sequences or with barcodes outside expected positions and lengths (14-16 nt) \nwere discarded. Reads, after barcode collapse: Number of reads remaining after merging \nsimilar barcodes. Collapsing merges counts into parent barcodes but does not discard reads. \nReads, after barcode filtering: Number of reads associated with barcodes that passed \nabundance and sequence filters. For the fish samples, barcodes were excluded from a fish if \nthey appeared <3 times in that fish. For the injected library, barcodes with <2 reads in the \ninjected library were excluded. Additionally, barcodes too similar (Levenshtein distance ≤2) to \nthe conserved non-barcode regions of the injected plasmid were removed. Barcodes, after \nextraction: Number of unique barcode sequences identified after extraction, before any quality \nfiltering or collapsing. Barcodes, after collapsing: Number of unique barcode sequences after \ncollapsing. Barcodes within Levenshtein distance of 1, assumed to result from PCR or \nsequencing errors, were merged into their most abundant neighbor (parent barcode), with read \ncounts combined. Barcodes, after filtering: Number of unique barcodes retained after applying \nthe abundance and sequence filters described above.  \nSample Reads, \n after \ndemultipl\nexing \nReads, \nafter \nbarcode \nextraction \nReads, \nafter \nbarcode \ncollapse \nReads, \nafter \nbarcode \nfiltering \nBarcodes\n, after \nextractio\nn \nBarcodes, \nafter \ncollapsin\ng \nBarcode\ns, after \nfiltering \nFish 1 604707 568573 568573 563410 8679 5933 1553 \nFish 2 572589 528807 528807 524074 9234 5797 1761 \nFish 3 567819 558715 558715 555185 6180 4699 1861 \nFish 4 538318 525189 525189 521701 5600 4330 1524 \nFish 5 643730 588909 588909 583426 10301 6749 1989 \nFish 6 598294 573249 573249 568572 8574 5700 1682 \nFish 7 598419 574577 574577 570243 7981 5383 1682 \nFish 8 634168 608624 608624 603129 10129 6623 1792 \nFish 9 584056 574166 574166 570225 6622 4728 1489 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nFish 10 602081 553224 553224 548768 8974 5666 1810 \nFish 11 552802 521451 521451 517234 7316 4937 1378 \nFish 12 568603 545314 545314 540172 9187 6118 1592 \nInjected \nlibrary \n5822820 5499123 5499123 3018743 4176867 3670283 1190778 \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nAppendix data S1: barcoded AttB-HS4-15N-nrUAS-GFP-CAAX plasmid sequence \nMap: AttB HS4 insulator DNA barcode 4xnrUAS GFP-CAAX \n \nAATACTCATACTCTTCCTTTTTCAATATTATTGAAGCATTTATCAGGGTTATTGTCTCATGAG\nCGGATACATATTTGAATGTATTTAGAAAAATAAACAAATAGGGGTTCCGCGCACATTTCCCC\nGAAAAGTGCCAGATACCTGAAACAAAACCCATCGTACGGCCAAGGAAGTCTCCAATAACTG\nTGATCCACCACAAGCGCCAGGGTTTTCCCAGTCACGACGTTGTAAAACGACGGCCAGTCA\nTGCATAATCCGCACGCATCTGGAATAAGGAAGTGCCATTCCGCCTGACCTCTCGAAGCCG\nCGGTGCGGGTGCCAGGGCGTGCCCTTGGGCTCCCCGGGCGCGTACTCCACCTCACCCAT\nCGAGCTCACGGGGACAGCCCCCTCCCAAAGCCCCCAGGGANNNNNNNNNNNNNNNCACG\nCTAGCTGTAATTACGTCCCTCCCCCGCTAGGGGGCAGCAGCGAGCCGCCCGGGGCTCCG\nCTCCGGTCCGGCGCTCCCCCCGCATCCCCGAGCCGGCAGCGTGCGGGGACAGCCCGGG\nCACGGGGAAGGTGGCACGGGATCGCTTTCCTCTGAACGCTTCTCGCTGCTCTTTGAGCCT\nGCAGACACCTGGGGGGATACGGGGAAAAAGCTTTAGGCTGAAAGAGAGATTTAGAATGAC\nAGGCGCGCCACTAGTCGGTGGCTTCTAATCCGTGAGTCCTAGCGGGTGACAGCCCTCCGT\nCTTCACAGGCGGAGGAGAGTCTTCCGTAGGGTTCCTCGGAGTACTGTCCTCCGACGCGTG\nCAAGGGTCGACTCTAGAGGGTATATAATGGATCCCATCGCGTCTCAGCCTCACTTTGAGCT\nCCTCCACACGCCACCATGGTTAGTAAAGGTGAGGAGCTGTTTACAGGTGTCGTGCCGATT\nCTCGTGGAACTTGACGGCGATGTAAATGGGCATAAATTCAGCGTATCTGGGGAAGGTGAG\nGGCGACGCAACTTACGGTAAACTGACCCTCAAGTTCATATGTACTACAGGGAAACTGCCTG\nTGCCGTGGCCTACTCTGGTAACAACTTTGACGTATGGCGTCCAATGTTTTAGCCGATATCC\nCGATCACATGAAACAACACGATTTCTTTAAATCAGCCATGCCTGAAGGATATGTGCAAGAA\nCGAACCATTTTCTTCAAAGACGATGGCAATTATAAAACCCGTGCAGAGGTTAAGTTTGAGG\nGCGATACACTCGTTAATCGGATCGAGCTGAAAGGAATAGACTTTAAGGAAGACGGCAATAT\nTCTGGGGCATAAACTGGAGTATAATTACAATTCACACAATGTCTACATCATGGCAGATAAGC\nAGAAGAACGGGATTAAAGTCAATTTCAAGATTAGACACAACATCGAAGACGGCTCCGTTCA\nACTCGCGGATCATTATCAGCAAAATACGCCCATCGGTGATGGCCCCGTTCTGCTCCCAGAT\nAACCACTATTTGAGCACGCAGAGCGCACTGTCAAAGGACCCTAATGAGAAAAGAGATCATA\nTGGTGCTCCTTGAGTTTGTTACAGCAGCTGGGATCACATTGGGGATGGATGAACTTTACAA\nAAAGCTGAACCCTCCTGATGAGAGTGGCCCCGGCTGCATGAGCTGCAAGTGTGTGCTCTC\nCTAAGATCCAGACATGATAAGATACATTGATGAGTTTGGACAAACCACAACTAGAATGCAGT\nGAAAAAAATGCTTTATTTGTGAAATTTGTGATGCTATTGCTTTATTTGTAACCATTATAAGCT\nGCAATAAACAAGTTAACAACAACAATTGCATTCATTTTATGTTTCAGGTTCAGGGGGAGGTG\nTGGGAGGTTTTTTAAAGGCTAGGTGGAGGCTCAGTGATGATAAGTCTGCGATGGTGGATG\nCATGTGTCATGGTCATAGCTGTTTCCTGTGTGAAATTGTTATCCGCTCAGAGGGCACAATC\nCTATTCCGCGCTATCCGACAATCTCCAAGACATTAGGTGGAGTTCAGTTCGGCGTATGGCA\nTATGTCGCTGGAAAGAACATGTGAGCAAAAGGCCAGCAAAAGGCCAGGAACCGTAAAAAG\nGCCGCGTTGCTGGCGTTTTTCCATAGGCTCCGCCCCCCTGACGAGCATCACAAAAATCGA\nCGCTCAAGTCAGAGGTGGCGAAACCCGACAGGACTATAAAGATACCAGGCGTTTCCCCCT\nGGAAGCTCCCTCGTGCGCTCTCCTGTTCCGACCCTGCCGCTTACCGGATACCTGTCCGCC\nTTTCTCCCTTCGGGAAGCGTGGCGCTTTCTCATAGCTCACGCTGTAGGTATCTCAGTTCGG\nTGTAGGTCGTTCGCTCCAAGCTGGGCTGTGTGCACGAACCCCCCGTTCAGCCCGACCGCT\nGCGCCTTATCCGGTAACTATCGTCTTGAGTCCAACCCGGTAAGACACGACTTATCGCCACT\nGGCAGCAGCCACTGGTAACAGGATTAGCAGAGCGAGGTATGTAGGCGGTGCTACAGAGTT\nCTTGAAGTGGTGGCCTAACTACGGCTACACTAGAAGAACAGTATTTGGTATCTGCGCTCTG\nCTGAAGCCAGTTACCTTCGGAAAAAGAGTTGGTAGCTCTTGATCCGGCAAACAAACCACCG\nCTGGTAGCGGTGGTTTTTTTGTTTGCAAGCAGCAGATTACGCGCAGAAAAAAAGGATCTCA\nAGAAGATCCTTTGATCTTTTCTACGGGGTCTGACGCTCTATTCAACAAAGCCGCCGTCCCG\nTCAAGTCAGCGTAAATGGGTAGGGGGCTTCAAATCGTCCTCGTGATACCAATTCGGAGCCT\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nGCTTTTTTGTACAAACTTGTTGATAATGGCAATTCAAGGATCTTCACCTAGATCCTTTTAAAT\nTAAAAATGAAGTTTTAAATCAATCTAAAGTATATATGAGTAAACTTGGTCTGACAGTTACCAA\nTGCTTAATCAGTGAGGCACCTATCTCAGCGATCTGTCTATTTCGTTCATCCATAGTTGCCTG\nACTCCCCGTCGTGTAGATAACTACGATACGGGAGGGCTTACCATCTGGCCCCAGTGCTGC\nAATGATACCGCGAGAGCCACGCTCACCGGCTCCAGATTTATCAGCAATAAACCAGCCAGC\nCGGAAGGGCCGAGCGCAGAAGTGGTCCTGCAACTTTATCCGCCTCCATCCAGTCTATTAA\nTTGTTGCCGGGAAGCTAGAGTAAGTAGTTCGCCAGTTAATAGTTTGCGCAACGTTGTTGCC\nATTGCTACAGGCATCGTGGTGTCACGCTCGTCGTTTGGTATGGCTTCATTCAGCTCCGGTT\nCCCAACGATCAAGGCGAGTTACATGATCCCCCATGTTGTGCAAAAAAGCGGTTAGCTCCTT\nCGGTCCTCCGATCGTTGTCAGAAGTAAGTTGGCCGCAGTGTTATCACTCATGGTTATGGCA\nGCACTGCATAATTCTCTTACTGTCATGCCATCCGTAAGATGCTTTTCTGTGACTGGTGAGTA\nCTCAACCAAGTCATTCTGAGAATAGTGTATGCGGCGACCGAGTTGCTCTTGCCCGGCGTC\nAATACGGGATAATACCGCGCCACATAGCAGAACTTTAAAAGTGCTCATCATTGGAAAACGT\nTCTTCGGGGCGAAAACTCTCAAGGATCTTACCGCTGTTGAGATCCAGTTCGATGTAACCCA\nCTCGTGCACCCAACTGATCTTCAGCATCTTTTACTTTCACCAGCGTTTCTGGGTGAGCAAAA\nACAGGAAGGCAAAATGCCGCAAAAAAGGGAATAAGGGCGACACGGAAATGTTG \n \n \nAppendix data S2: AttB-HS4-nrUAS-GFP-CAAX plasmid sequence \nCGACGTTGTAAAACGACGGCCAGTCATGCATAATCCGCACGCATCTGGAATAAGGAAGTG\nCCATTCCGCCTGACCTCTCGAAGCCGCGGTGCGGGTGCCAGGGCGTGCCCTTGGGCTCC\nCCGGGCGCGTACTCCACCTCACCCATCGAGCTCACGGGGACAGCCCCCTCCCAAAGCCC\nCCAGGGATGTAATTACGTCCCTCCCCCGCTAGGGGGCAGCAGCGAGCCGCCCGGGGCTC\nCGCTCCGGTCCGGCGCTCCCCCCGCATCCCCGAGCCGGCAGCGTGCGGGGACAGCCCG\nGGCACGGGGAAGGTGGCACGGGATCGCTTTCCTCTGAACGCTTCTCGCTGCTCTTTGAGC\nCTGCAGACACCTGGGGGGATACGGGGAAAAAGCTTTAGGCTGAAAGAGAGATTTAGAATG\nACAGGCGCGCCACTAGTCGGTGGCTTCTAATCCGTGAGTCCTAGCGGGTGACAGCCCTCC\nGTCTTCACAGGCGGAGGAGAGTCTTCCGTAGGGTTCCTCGGAGTACTGTCCTCCGACGCG\nTGCAAGGGTCGACTCTAGAGGGTATATAATGGATCCCATCGCGTCTCAGCCTCACTTTGAG\nCTCCTCCACACGCCACCATGGTTAGTAAAGGTGAGGAGCTGTTTACAGGTGTCGTGCCGA\nTTCTCGTGGAACTTGACGGCGATGTAAATGGGCATAAATTCAGCGTATCTGGGGAAGGTGA\nGGGCGACGCAACTTACGGTAAACTGACCCTCAAGTTCATATGTACTACAGGGAAACTGCCT\nGTGCCGTGGCCTACTCTGGTAACAACTTTGACGTATGGCGTCCAATGTTTTAGCCGATATC\nCCGATCACATGAAACAACACGATTTCTTTAAATCAGCCATGCCTGAAGGATATGTGCAAGA\nACGAACCATTTTCTTCAAAGACGATGGCAATTATAAAACCCGTGCAGAGGTTAAGTTTGAG\nGGCGATACACTCGTTAATCGGATCGAGCTGAAAGGAATAGACTTTAAGGAAGACGGCAATA\nTTCTGGGGCATAAACTGGAGTATAATTACAATTCACACAATGTCTACATCATGGCAGATAAG\nCAGAAGAACGGGATTAAAGTCAATTTCAAGATTAGACACAACATCGAAGACGGCTCCGTTC\nAACTCGCGGATCATTATCAGCAAAATACGCCCATCGGTGATGGCCCCGTTCTGCTCCCAGA\nTAACCACTATTTGAGCACGCAGAGCGCACTGTCAAAGGACCCTAATGAGAAAAGAGATCAT\nATGGTGCTCCTTGAGTTTGTTACAGCAGCTGGGATCACATTGGGGATGGATGAACTTTACA\nAAAAGCTGAACCCTCCTGATGAGAGTGGCCCCGGCTGCATGAGCTGCAAGTGTGTGCTCT\nCCTAAGATCCAGACATGATAAGATACATTGATGAGTTTGGACAAACCACAACTAGAATGCA\nGTGAAAAAAATGCTTTATTTGTGAAATTTGTGATGCTATTGCTTTATTTGTAACCATTATAAG\nCTGCAATAAACAAGTTAACAACAACAATTGCATTCATTTTATGTTTCAGGTTCAGGGGGAGG\nTGTGGGAGGTTTTTTAAAGGCTAGGTGGAGGCTCAGTGATGATAAGTCTGCGATGGTGGA\nTGCATGTGTCATGGTCATAGCTGTTTCCTGTGTGAAATTGTTATCCGCTCAGAGGGCACAA\nTCCTATTCCGCGCTATCCGACAATCTCCAAGACATTAGGTGGAGTTCAGTTCGGCGTATGG\nCATATGTCGCTGGAAAGAACATGTGAGCAAAAGGCCAGCAAAAGGCCAGGAACCGTAAAA\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nAGGCCGCGTTGCTGGCGTTTTTCCATAGGCTCCGCCCCCCTGACGAGCATCACAAAAATC\nGACGCTCAAGTCAGAGGTGGCGAAACCCGACAGGACTATAAAGATACCAGGCGTTTCCCC\nCTGGAAGCTCCCTCGTGCGCTCTCCTGTTCCGACCCTGCCGCTTACCGGATACCTGTCCG\nCCTTTCTCCCTTCGGGAAGCGTGGCGCTTTCTCATAGCTCACGCTGTAGGTATCTCAGTTC\nGGTGTAGGTCGTTCGCTCCAAGCTGGGCTGTGTGCACGAACCCCCCGTTCAGCCCGACC\nGCTGCGCCTTATCCGGTAACTATCGTCTTGAGTCCAACCCGGTAAGACACGACTTATCGCC\nACTGGCAGCAGCCACTGGTAACAGGATTAGCAGAGCGAGGTATGTAGGCGGTGCTACAGA\nGTTCTTGAAGTGGTGGCCTAACTACGGCTACACTAGAAGAACAGTATTTGGTATCTGCGCT\nCTGCTGAAGCCAGTTACCTTCGGAAAAAGAGTTGGTAGCTCTTGATCCGGCAAACAAACCA\nCCGCTGGTAGCGGTGGTTTTTTTGTTTGCAAGCAGCAGATTACGCGCAGAAAAAAAGGATC\nTCAAGAAGATCCTTTGATCTTTTCTACGGGGTCTGACGCTCTATTCAACAAAGCCGCCGTC\nCCGTCAAGTCAGCGTAAATGGGTAGGGGGCTTCAAATCGTCCTCGTGATACCAATTCGGA\nGCCTGCTTTTTTGTACAAACTTGTTGATAATGGCAATTCAAGGATCTTCACCTAGATCCTTTT\nAAATTAAAAATGAAGTTTTAAATCAATCTAAAGTATATATGAGTAAACTTGGTCTGACAGTTA\nCCAATGCTTAATCAGTGAGGCACCTATCTCAGCGATCTGTCTATTTCGTTCATCCATAGTTG\nCCTGACTCCCCGTCGTGTAGATAACTACGATACGGGAGGGCTTACCATCTGGCCCCAGTG\nCTGCAATGATACCGCGAGAGCCACGCTCACCGGCTCCAGATTTATCAGCAATAAACCAGC\nCAGCCGGAAGGGCCGAGCGCAGAAGTGGTCCTGCAACTTTATCCGCCTCCATCCAGTCTA\nTTAATTGTTGCCGGGAAGCTAGAGTAAGTAGTTCGCCAGTTAATAGTTTGCGCAACGTTGT\nTGCCATTGCTACAGGCATCGTGGTGTCACGCTCGTCGTTTGGTATGGCTTCATTCAGCTCC\nGGTTCCCAACGATCAAGGCGAGTTACATGATCCCCCATGTTGTGCAAAAAAGCGGTTAGCT\nCCTTCGGTCCTCCGATCGTTGTCAGAAGTAAGTTGGCCGCAGTGTTATCACTCATGGTTAT\nGGCAGCACTGCATAATTCTCTTACTGTCATGCCATCCGTAAGATGCTTTTCTGTGACTGGT\nGAGTACTCAACCAAGTCATTCTGAGAATAGTGTATGCGGCGACCGAGTTGCTCTTGCCCG\nGCGTCAATACGGGATAATACCGCGCCACATAGCAGAACTTTAAAAGTGCTCATCATTGGAA\nAACGTTCTTCGGGGCGAAAACTCTCAAGGATCTTACCGCTGTTGAGATCCAGTTCGATGTA\nACCCACTCGTGCACCCAACTGATCTTCAGCATCTTTTACTTTCACCAGCGTTTCTGGGTGA\nGCAAAAACAGGAAGGCAAAATGCCGCAAAAAAGGGAATAAGGGCGACACGGAAATGTTGA\nATACTCATACTCTTCCTTTTTCAATATTATTGAAGCATTTATCAGGGTTATTGTCTCATGAGC\nGGATACATATTTGAATGTATTTAGAAAAATAAACAAATAGGGGTTCCGCGCACATTTCCCCG\nAAAAGTGCCAGATACCTGAAACAAAACCCATCGTACGGCCAAGGAAGTCTCCAATAACTGT\nGATCCACCACAAGCGCCAGGGTTTTCCCAGTCA \n \n \nAppendix data S3: AttB-HS4-nrUAS-mScarlet-CAAX plasmid sequence \nCGACGTTGTAAAACGACGGCCAGTCATGCATAATCCGCACGCATCTGGAATAAGGAAGTG\nCCATTCCGCCTGACCTCTCGAAGCCGCGGTGCGGGTGCCAGGGCGTGCCCTTGGGCTCC\nCCGGGCGCGTACTCCACCTCACCCATCGAGCTCACGGGGACAGCCCCCTCCCAAAGCCC\nCCAGGGATGTAATTACGTCCCTCCCCCGCTAGGGGGCAGCAGCGAGCCGCCCGGGGCTC\nCGCTCCGGTCCGGCGCTCCCCCCGCATCCCCGAGCCGGCAGCGTGCGGGGACAGCCCG\nGGCACGGGGAAGGTGGCACGGGATCGCTTTCCTCTGAACGCTTCTCGCTGCTCTTTGAGC\nCTGCAGACACCTGGGGGGATACGGGGAAAAAGCTTTAGGCTGAAAGAGAGATTTAGAATG\nACAGGCGCGCCACTAGTCGGTGGCTTCTAATCCGTGAGTCCTAGCGGGTGACAGCCCTCC\nGTCTTCACAGGCGGAGGAGAGTCTTCCGTAGGGTTCCTCGGAGTACTGTCCTCCGACGCG\nTGCAAGGGTCGACTCTAGAGGGTATATAATGGATCCCATCGCGTCTCAGCCTCACTTTGAG\nCTCCTCCACACGCCACCATGGTGAGCAAGGGCGAGGCAGTGATCAAGGAGTTCATGCGGT\nTCAAGGTGCACATGGAGGGCTCCATGAACGGCCACGAGTTCGAGATCGAGGGCGAGGGC\nGAGGGCCGCCCCTACGAGGGCACCCAGACCGCCAAGCTGAAGGTGACCAAGGGTGGCC\nCCCTGCCCTTCTCCTGGGACATCCTGTCCCCTCAGTTCATGTACGGCTCCAGGGCCTTCA\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nCCAAGCACCCCGCCGACATCCCCGACTACTATAAGCAGTCCTTCCCCGAGGGCTTCAAGT\nGGGAGCGCGTGATGAACTTCGAGGACGGCGGCGCCGTGACCGTGACCCAGGACACCTCC\nCTGGAGGACGGCACCCTGATCTACAAGGTGAAGCTCCGCGGCACCAACTTCCCTCCTGAC\nGGCCCCGTAATGCAGAAGAAGACAATGGGCTGGGAAGCGTCCACCGAGCGGTTGTACCC\nCGAGGACGGCGTGCTGAAGGGCGACATTAAGATGGCCCTGCGCCTGAAGGACGGCGGCC\nGATACCTGGCGGACTTCAAGACCACCTACAAGGCCAAGAAGCCCGTGCAGATGCCCGGC\nGCCTACAACGTGGACCGCAAGTTGGACATCACCTCCCACAACGAGGACTACACCGTGGTG\nGAACAGTACGAACGCTCCGAGGGCCGCCACTCCACCGGCGGCATGGACGAGCTGTACAA\nGAAGCTGAACCCTCCTGATGAGAGTGGCCCCGGCTGCATGAGCTGCAAGTGTGTGCTCTC\nCTAAGATCCAGACATGATAAGATACATTGATGAGTTTGGACAAACCACAACTAGAATGCAGT\nGAAAAAAATGCTTTATTTGTGAAATTTGTGATGCTATTGCTTTATTTGTAACCATTATAAGCT\nGCAATAAACAAGTTAACAACAACAATTGCATTCATTTTATGTTTCAGGTTCAGGGGGAGGTG\nTGGGAGGTTTTTTAAAGGCTAGGTGGAGGCTCAGTGATGATAAGTCTGCGATGGTGGATG\nCATGTGTCATGGTCATAGCTGTTTCCTGTGTGAAATTGTTATCCGCTCAGAGGGCACAATC\nCTATTCCGCGCTATCCGACAATCTCCAAGACATTAGGTGGAGTTCAGTTCGGCGTATGGCA\nTATGTCGCTGGAAAGAACATGTGAGCAAAAGGCCAGCAAAAGGCCAGGAACCGTAAAAAG\nGCCGCGTTGCTGGCGTTTTTCCATAGGCTCCGCCCCCCTGACGAGCATCACAAAAATCGA\nCGCTCAAGTCAGAGGTGGCGAAACCCGACAGGACTATAAAGATACCAGGCGTTTCCCCCT\nGGAAGCTCCCTCGTGCGCTCTCCTGTTCCGACCCTGCCGCTTACCGGATACCTGTCCGCC\nTTTCTCCCTTCGGGAAGCGTGGCGCTTTCTCATAGCTCACGCTGTAGGTATCTCAGTTCGG\nTGTAGGTCGTTCGCTCCAAGCTGGGCTGTGTGCACGAACCCCCCGTTCAGCCCGACCGCT\nGCGCCTTATCCGGTAACTATCGTCTTGAGTCCAACCCGGTAAGACACGACTTATCGCCACT\nGGCAGCAGCCACTGGTAACAGGATTAGCAGAGCGAGGTATGTAGGCGGTGCTACAGAGTT\nCTTGAAGTGGTGGCCTAACTACGGCTACACTAGAAGAACAGTATTTGGTATCTGCGCTCTG\nCTGAAGCCAGTTACCTTCGGAAAAAGAGTTGGTAGCTCTTGATCCGGCAAACAAACCACCG\nCTGGTAGCGGTGGTTTTTTTGTTTGCAAGCAGCAGATTACGCGCAGAAAAAAAGGATCTCA\nAGAAGATCCTTTGATCTTTTCTACGGGGTCTGACGCTCTATTCAACAAAGCCGCCGTCCCG\nTCAAGTCAGCGTAAATGGGTAGGGGGCTTCAAATCGTCCTCGTGATACCAATTCGGAGCCT\nGCTTTTTTGTACAAACTTGTTGATAATGGCAATTCAAGGATCTTCACCTAGATCCTTTTAAAT\nTAAAAATGAAGTTTTAAATCAATCTAAAGTATATATGAGTAAACTTGGTCTGACAGTTACCAA\nTGCTTAATCAGTGAGGCACCTATCTCAGCGATCTGTCTATTTCGTTCATCCATAGTTGCCTG\nACTCCCCGTCGTGTAGATAACTACGATACGGGAGGGCTTACCATCTGGCCCCAGTGCTGC\nAATGATACCGCGAGAGCCACGCTCACCGGCTCCAGATTTATCAGCAATAAACCAGCCAGC\nCGGAAGGGCCGAGCGCAGAAGTGGTCCTGCAACTTTATCCGCCTCCATCCAGTCTATTAA\nTTGTTGCCGGGAAGCTAGAGTAAGTAGTTCGCCAGTTAATAGTTTGCGCAACGTTGTTGCC\nATTGCTACAGGCATCGTGGTGTCACGCTCGTCGTTTGGTATGGCTTCATTCAGCTCCGGTT\nCCCAACGATCAAGGCGAGTTACATGATCCCCCATGTTGTGCAAAAAAGCGGTTAGCTCCTT\nCGGTCCTCCGATCGTTGTCAGAAGTAAGTTGGCCGCAGTGTTATCACTCATGGTTATGGCA\nGCACTGCATAATTCTCTTACTGTCATGCCATCCGTAAGATGCTTTTCTGTGACTGGTGAGTA\nCTCAACCAAGTCATTCTGAGAATAGTGTATGCGGCGACCGAGTTGCTCTTGCCCGGCGTC\nAATACGGGATAATACCGCGCCACATAGCAGAACTTTAAAAGTGCTCATCATTGGAAAACGT\nTCTTCGGGGCGAAAACTCTCAAGGATCTTACCGCTGTTGAGATCCAGTTCGATGTAACCCA\nCTCGTGCACCCAACTGATCTTCAGCATCTTTTACTTTCACCAGCGTTTCTGGGTGAGCAAAA\nACAGGAAGGCAAAATGCCGCAAAAAAGGGAATAAGGGCGACACGGAAATGTTGAATACTC\nATACTCTTCCTTTTTCAATATTATTGAAGCATTTATCAGGGTTATTGTCTCATGAGCGGATAC\nATATTTGAATGTATTTAGAAAAATAAACAAATAGGGGTTCCGCGCACATTTCCCCGAAAAGT\nGCCAGATACCTGAAACAAAACCCATCGTACGGCCAAGGAAGTCTCCAATAACTGTGATCCA\nCCACAAGCGCCAGGGTTTTCCCAGTCA \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nAppendix data S4: Sequences of key primers used for library generation and sequencing \nPrimer sequence Primer name Use \nGTTGATCATACATTGGCACGGCTAGCTGTAAT\nTACGTCCCTCCCCCGCTA \nF_HS4_NheI 2-cycle PCR to add \nrandom 15xN barcodes in \nthe middle of the HS4 \nelement for the plasmid \nlibrary. An NheI restriction \nsite is also added with \nthese primers to enable \nspecific sticky-end \nrestriction-ligation for \nrecircularization of the \nplasmid.  \nAGTCAAGTGGAATACTGCTAGCGTGNNNNNN\nNNNNNNNNNTCCCTGGGGGCTTTGGGAGG \nR_Add15N_HS4_Nh\neI \nCTTACTCATACATTGGCACGGC F_amp_HS4 Amplification of the \nbarcoded linearized whole-\nplasmid amplicons  \nAGTCAAGTGGAATACTGCTAGCG R_amp_HS4 \nACAACCCGACAGCCTACGTCAC F_Chr24pIGLET Specific amplification of \nintegrated barcodes from \ngenomic extracts of mosaic \nzebrafish, generating a 652 \nbp amplicon library  \nGAGAAGCGTTCAGAGGAAAGCGATC R_HS4 \nTCGTCGGCAGCGTCAGATGTGTATAAGAGAC\nAGTGGAGATCACTTCATTCTATTTTCCCT \nF_Chr24_illumread Generation of 325 bp \namplicon library of fish-\nrecovered integrated \nbarcodes for direct Illumina \nsequencing, based on \namplification of the 652 bp \namplicon library and \naddition of Illumina \noverhangs and sample-\nspecific 5-nt barcode for \ndemultiplexing \nGTCTCGTGGGCTCGGAGATGTGTATAAGAGA\nCAGNNNNNTAGCGGGGGAGGGACGTAATT \nR_HS4_FishX_illumr\nead \nTCGTCGGCAGCGTCAGATGTGTATAAGAGAC\nAGACGGGGACAGCCCCCTCCCAAAG \nF_HS4_illumread Generation of 336 bp \namplicon library of \nbarcodes from the injected \nplasmid library for direct \nIllumina sequencing GTCTCGTGGGCTCGGAGATGTGTATAAGAGA\nCAGCAGCCTAAAGCTTTTTCCCCGTATCC \nR_HS4_illumread \n \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nReferences \n \n1.  M. Kuhn, A. J. Santinha, R. J. Platt, Moving from in vitro to in vivo CRISPR screens. Gene \nand Genome Editing 2, 100008 (2021). \n2.  F. Vincent, et al., Phenotypic drug discovery: recent successes, lessons learned and new \ndirections. Nat. Rev. Drug Discov. 21, 899–914 (2022). \n3.  J. G. Moffat, F. Vincent, J. A. Lee, J. Eder, M. Prunotto, Opportunities and challenges in \nphenotypic drug discovery: an industry perspective. Nat. Rev. Drug Discov. 16, 531–543 \n(2017). \n4.  P. Horvath, et al., Screening out irrelevant cell-based models of disease. Nat. Rev. Drug \nDiscov. 15, 751–769 (2016). \n5.  J. Acosta, et al., Multiplexed in vivo base editing identifies functional gene-variant-context \ninteractions. bioRxivorg 2025.02.23.639770 (2025). \n6.  J. W. Scannell, J. Bosley, When quality beats quantity: Decision theory, drug discovery, \nand the reproducibility crisis. PLoS One 11, e0147215 (2016). \n7.  J. Yang, et al., Solaris: a panel of bright and sensitive hybrid voltage indicators for imaging \nmembrane potential in cultured neurons. bioRxiv 2024.02.02.578569 (2024). \n8.  M. Kannan, G. Vasan, V. A. Pieribone, Optimizing Strategies for Developing Genetically \nEncoded Voltage Indicators. Front. Cell. Neurosci. 13, 53 (2019). \n9.  O. A. Shemesh, et al., Precision Calcium Imaging of Dense Neural Populations via a Cell-\nBody-Targeted Calcium Indicator. Neuron 107, 470–486.e11 (2020). \n10.  Y. Chen, et al., Soma-targeted imaging of neural circuits by ribosome tethering. Neuron \n107, 454–469.e6 (2020). \n11.  A. E. Palmer, et al., Ca2+ indicators based on computationally redesigned calmodulin-\npeptide pairs. Chem. Biol. 13, 521–530 (2006). \n12.  Y. Yang, et al., Improved calcium sensor GCaMP-X overcomes the calcium channel \nperturbations induced by the calmodulin in GCaMP. Nat. Commun. 9, 1504 (2018). \n13.  H.-W. Yeh, T. Wu, M. Chen, H.-W. Ai, Identification of factors complicating \nbioluminescence imaging. Biochemistry 58, 1689–1697 (2019). \n14.  M. Haeussler, CRISPR off-targets: a question of context. Cell Biol. Toxicol. 36, 5–9 (2020). \n15.  E. Zuo, et al., Cytosine base editor generates substantial off-target single-nucleotide \nvariants in mouse embryos. Science 364, 289–292 (2019). \n16.  W. L. Chew, et al., A multifunctional AAV-CRISPR-Cas9 and its host response. Nat. \nMethods 13, 868–874 (2016). \n17.  M. Schmidt-Supprian, K. Rajewsky, Vagaries of conditional gene targeting. Nat. Immunol. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\n8, 665–668 (2007). \n18.  L. Tian, et al., Imaging neural activity in worms, flies and mice with improved GCaMP \ncalcium indicators. Nat. Methods 6, 875–881 (2009). \n19.  B. E. Maimon, et al., Optogenetic Peripheral Nerve Immunogenicity. Sci. Rep. 8, 14076 \n(2018). \n20.  T. Miyashita, Y. R. Shao, J. Chung, O. Pourzia, D. E. Feldman, Long-term \nchannelrhodopsin-2 (ChR2) expression can induce abnormal axonal morphology and \ntargeting in cerebral cortex. Front. Neural Circuits 7, 8 (2013). \n21.  M. Borch Jensen, A. Marblestone, In vivo Pooled Screening: A Scalable Tool to Study the \nComplexity of Aging and Age-Related Disease. Front Aging 2, 714926 (2021). \n22.  X. Jin, et al., In vivo Perturb-Seq reveals neuronal and glial abnormalities associated with \nautism risk genes. Science 370 (2020). \n23.  A. J. Santinha, et al., Transcriptional linkage analysis with in vivo AAV-Perturb-seq. Nature \n622, 367–375 (2023). \n24.  X. Zheng, et al., Massively parallel in vivo Perturb-seq reveals cell-type-specific \ntranscriptional networks in cortical development. Cell 187, 3236–3248.e21 (2024). \n25.  A. J. Santinha, A. Strano, R. J. Platt, Methods and applications of in vivo CRISPR \nscreening. Nat. Rev. Genet. 26, 702–718 (2025). \n26.  M. H. Wertz, et al., Genome-wide in vivo CNS screening identifies genes that modify CNS \nneuronal survival and mHTT toxicity. Neuron 106, 76–89.e8 (2020). \n27.  R. A. Saunders, et al., Perturb-Multimodal: A platform for pooled genetic screens with \nimaging and sequencing in intact mammalian tissue. Cell 188, 4790–4809.e22 (2025). \n28.  C. J. Walkey, et al., A comprehensive atlas of AAV tropism in the mouse. Mol. Ther. 33, \n1282–1299 (2025). \n29.  A. Pupo, et al., AAV vectors: The Rubik’s cube of human gene therapy. Mol. Ther. 30, \n3515–3541 (2022). \n30.  M. Mietzsch, F. Broecker, A. Reinhardt, P. H. Seeberger, R. Heilbronn, Differential adeno-\nassociated virus serotype-specific interaction patterns with synthetic heparins and other \nglycans. J. Virol. 88, 2991–3003 (2014). \n31.  T. J. Gonzalez, et al., Cross-species evolution of a highly potent AAV variant for therapeutic \ngene transfer and genome editing. Nat. Commun. 13, 5947 (2022). \n32.  G. Kalamakis, R. J. Platt, CRISPR for neuroscientists. Neuron 111, 2282–2311 (2023). \n33.  R. T. Manguso, et al., In vivo CRISPR screening identifies Ptpn2 as a cancer \nimmunotherapy target. Nature 547, 413–418 (2017). \n34.  N. J. VanDusen, et al., Massively parallel in vivo CRISPR screening identifies RNF20/40 as \nepigenetic regulators of cardiomyocyte maturation. Nat. Commun. 12, 4442 (2021). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\n35.  R. D. Chow, et al., AAV-mediated direct in vivo CRISPR screen identifies functional \nsuppressors in glioblastoma. Nat. Neurosci. 20, 1329–1341 (2017). \n36.  T. J. Ruetz, et al., CRISPR-Cas9 screens reveal regulators of ageing in neural stem cells. \nNature 634, 1150–1159 (2024). \n37.  Z. C. Stevenson, et al., High-throughput library transgenesis in Caenorhabditis elegans via \nTransgenic Arrays Resulting in Diversity of Integrated Sequences (TARDIS). Elife 12 \n(2023). \n38.  D. J. Dickinson, A. M. Pani, J. K. Heppert, C. D. Higgins, B. Goldstein, Streamlined genome \nengineering with a self-excising drug selection cassette. Genetics 200, 1035–1049 (2015). \n39.  D. J. Dickinson, B. Goldstein, CRISPR-Based Methods for Caenorhabditis elegans \nGenome Engineering. Genetics 202, 885–901 (2016). \n40.  H. Liao, J. Wu, N. J. VanDusen, Y. Li, Y. Zheng, CRISPR-Cas9-mediated homology-\ndirected repair for precise gene editing. Mol. Ther. Nucleic Acids 35, 102344 (2024). \n41.  G. W. Stuart, J. V. McMurray, M. Westerfield, Replication, integration and stable germ-line \ntransmission of foreign sequences injected into early zebrafish embryos. Development 103, \n403–412 (1988). \n42.  A. Amsterdam, S. Lin, N. Hopkins, The Aequorea victoria green fluorescent protein can be \nused as a reporter in live zebrafish embryos. Dev. Biol. 171, 123–129 (1995). \n43.  P. Collas, Modulation of plasmid DNA methylation and expression in zebrafish embryos. \nNucleic Acids Res. 26, 4454–4461 (1998). \n44.  M. L. Suster, H. Kikuta, A. Urasaki, K. Asakawa, K. Kawakami, Transgenesis in zebrafish \nwith the tol2 transposon system. Methods Mol. Biol. 561, 41–63 (2009). \n45.  C. Mosimann, et al., Site-directed zebrafish transgenesis into single landing sites with the \nphiC31 integrase system: Zebrafish Transgenesis with phiC31. Dev. Dyn. 242, 949–963 \n(2013). \n46.  R. L. Lalonde, et al., pIGLET: Safe harbor landing sites for reproducible and efficient \ntransgenesis in zebrafish. Sci Adv 10, eadn6603 (2024). \n47.  K. A. Matreyek, J. J. Stephany, M. A. Chiasson, N. Hasle, D. M. Fowler, An improved \nplatform for functional assessment of large protein libraries in mammalian cells. Nucleic \nAcids Res. 48, e1 (2020). \n48.  C. M. Akitake, M. Macurak, M. E. Halpern, M. G. Goll, Transgenerational analysis of \ntranscriptional silencing in zebrafish. Dev. Biol. 352, 191–201 (2011). \n49.  O. Randlett, et al., Whole-brain activity mapping onto a zebrafish brain atlas. Nat. Methods \n12, 1039–1046 (2015). \n50.  M. B. Ahrens, M. B. Orger, D. N. Robson, J. M. Li, P. J. Keller, Whole-brain functional \nimaging at cellular resolution using light-sheet microscopy. Nat. Methods 10, 413–420 \n(2013). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\n51.  X. Chen, et al., Whole-brain light-sheet imaging data. Janelia Research Campus. \nhttps://doi.org/10.25378/JANELIA.7272617.V4. Deposited 7 February 2019. \n52.  T. Bondue, S. P. Berlingerio, L. van den Heuvel, E. Levtchenko, The zebrafish embryo as a \nmodel organism for testing mRNA-based therapeutics. Int. J. Mol. Sci. 24, 11224 (2023). \n53.  ZFIN Zebrafish Developmental Stages. Available at: https://zfin.org/zf_info/zfbook/stages/ \n[Accessed 2 November 2025]. \n54.  S. Y. Park, et al., Combinatorial protein barcodes enable self-correcting neuron tracing with \nnanoscale molecular context. bioRxivorg 2025.09.26.678648 (2025). \n55.  A. McKenna, et al., Whole-organism lineage tracing by combinatorial and cumulative \ngenome editing. Science 353, aaf7907 (2016). \n56.  J. Cotterell, M. Vila-Cejudo, L. Batlle-Morera, J. Sharpe, Endogenous CRISPR/Cas9 arrays \nfor scalable whole-organism lineage tracing. Development 147, dev184481 (2020). \n57.  B. Spanjaard, et al., Simultaneous lineage tracing and cell-type identification using \nCRISPR-Cas9-induced genetic scars. Nat. Biotechnol. 36, 469–473 (2018). \n58.  E. Raz, Primordial germ-cell development: the zebrafish perspective. Nat. Rev. Genet. 4, \n690–700 (2003). \n59.  B. E. Hew, et al., Directed evolution of hyperactive integrases for site specific insertion of \ntransgenes. Nucleic Acids Res. (2024). https://doi.org/10.1093/nar/gkae534. \n60.  D. A. Kane, C. B. Kimmel, The zebrafish midblastula transition. Development 119, 447–456 \n(1993). \n61.  W. Shoji, M. Sato-Maeda, Application of heat shock promoter in transgenic zebrafish. Dev. \nGrowth Differ. 50, 401–406 (2008). \n62.  S. S. Gerety, et al., An inducible transgene expression system for zebrafish and chick. \nDevelopment 140, 2235–2243 (2013). \n63.  A. Varady, et al., zHORSE as an optogenetic zebrafish strain for precise spatiotemporal \ncontrol over gene expression during development. Dev. Cell 60, 2825–2839.e4 (2025). \n64.  T.-Y. Choi, T.-I. Choi, Y.-R. Lee, S.-K. Choe, C.-H. Kim, Zebrafish as an animal model for \nbiomedical research. Exp. Mol. Med. 53, 310–317 (2021). \n65.  E. E. Patton, L. I. Zon, D. M. Langenau, Zebrafish disease models in drug discovery: from \npreclinical modelling to clinical trials. Nat. Rev. Drug Discov. 20, 611–628 (2021). \n66.  J.-M. Knapp, P. Chung, J. H. Simpson, Generating customized transgene landing sites and \nmulti-transgene arrays in Drosophila using phiC31 integrase. Genetics 199, 919–934 \n(2015). \n67.  J. Bischof, R. K. Maeda, M. Hediger, F. Karch, K. Basler, An optimized transgenesis \nsystem for Drosophila using germ-line-specific phiC31 integrases. Proc. Natl. Acad. Sci. U. \nS. A. 104, 3312–3317 (2007). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\n68.  Y. Bi, et al., ФC31 Integrase-Mediated Isolation and Characterization of Novel Safe Harbors \nfor Transgene Expression in the Pig Genome. Int. J. Mol. Sci. 19 (2018). \n69.  F.-J. Yang, et al., phiC31 integrase for recombination-mediated single-copy insertion and \ngenome manipulation in Caenorhabditis elegans. Genetics 220 (2022). \n70.  B. E. Low, V. Hosur, S. Lesbirel, M. V. Wiles, Efficient targeted transgenesis of large donor \nDNA into multiple mouse genetic backgrounds using bacteriophage Bxb1 integrase. Sci. \nRep. 12, 5424 (2022). \n71.  B. Tasic, et al., Site-specific integrase-mediated transgenesis in mice via pronuclear \ninjection. Proc. Natl. Acad. Sci. U. S. A. 108, 7902–7907 (2011). \n \n \n \n \n \n \n \n \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint \n\nAdditional information \n \nContact information: Edward Boyden (edboyden@mit.edu) \n \nCompeting interests: The authors declare no competing interests. \n \nData sharing plans: \nAll the raw imaging data and Illumina sequencing data associated with Fig. 2, 3, S1, S2, S3, S4, \nS5, S6, and Table S1, S2, S3, are available on DOI https://doi.org/10.5061/dryad.d2547d8h0. \nAll the code used to analyze the sequencing data is available on \nhttps://github.com/shaharbr/library_transgenesis. The full sequences of key plasmids and \nprimers used in this study are available in appendix data S1, S2, S3 and S4.  \nFunding information:  \nESB acknowledges, for funding, Lisa Yang, HHMI, NIH 1U01NS120820, NIH 1R01MH123977, \nNIH R01MH122971, and NIH R01DA029639. SB acknowledges funding from the Y. Eva Tan \nPostdoctoral Fellowship. \n \nSignificance statement: Genetic perturbations and molecular tools characterized in cell culture \nfrequently fail to translate in vivo, yet pooled screening in living animals faces critical limitations: \nthe high prevalence of multi-transgene cells confounds interpretation, viral packaging constrains \ntransgene size, and tropism introduces biases. We developed a library transgenesis method, \nimplemented in zebrafish, that overcomes these challenges by exploiting delayed site-specific \nintegration to create mosaic animals with >1,500 multi-kilobase transgenes integrated per \nanimal. In those library mosaics, ~99% of the cells express a single library member, thanks to \nthe mutual-exclusivity enforced by the site-specific integration mechanism. Library transgenesis \ncan transform each animal into hundreds of parallel experiments, enabling direct in vivo \nscreening of molecular tools and genetic perturbations in their native physiological contexts.  \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted January 31, 2026. ; https://doi.org/10.64898/2026.01.30.702415doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}