CLONEID: A Framework for Longitudinal Integration of Phenotypic and Genotypic Data to Monitor and Steer Subclonal Dynamics

A longitudinal data framework for context-specific genotype-to-phenotype mapping | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results A longitudinal data framework for context-specific genotype-to-phenotype mapping View ORCID Profile Thomas Veith , View ORCID Profile Richard Beck , View ORCID Profile Vural Tagal , View ORCID Profile Tao Li , View ORCID Profile Saeed Alahmari , View ORCID Profile Jackson Cole , Daniel Hannaby , John Kyei , View ORCID Profile Xiaoqing Yu , View ORCID Profile Konstantin Maksin , View ORCID Profile Andrew Schultz , View ORCID Profile HoJoon Lee , View ORCID Profile Aaron Diaz , Janine Lupo , View ORCID Profile Issam ElNaqa , View ORCID Profile Steven Eschrich , View ORCID Profile Hanlee P. Ji , View ORCID Profile Noemi Andor doi: https://doi.org/10.1101/2025.05.07.652202 Thomas Veith 2 Department of Integrated Mathematical Oncology, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Thomas Veith Richard Beck 2 Department of Integrated Mathematical Oncology, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Richard Beck Vural Tagal 2 Department of Integrated Mathematical Oncology, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Vural Tagal Tao Li 2 Department of Integrated Mathematical Oncology, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Tao Li Saeed Alahmari 1 Department of Computer Science, Najran University , Najran 66462, Saudi Arabia Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Saeed Alahmari Jackson Cole 2 Department of Integrated Mathematical Oncology, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jackson Cole Daniel Hannaby Find this author on Google Scholar Find this author on PubMed Search for this author on this site John Kyei 2 Department of Integrated Mathematical Oncology, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Xiaoqing Yu 3 Department of Biostatistics and Bioinformatics, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Xiaoqing Yu Konstantin Maksin 8 The University of Commerce and Services, Medical Department (WSHIU) , Poznan, Poland Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Konstantin Maksin Andrew Schultz 2 Department of Integrated Mathematical Oncology, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Andrew Schultz HoJoon Lee 5 Department of Medicine, Stanford University , CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for HoJoon Lee Aaron Diaz 6 Department of Epidemiology & Biostatistics, University of California San Francisco , CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Aaron Diaz Janine Lupo 1 Department of Computer Science, Najran University , Najran 66462, Saudi Arabia Find this author on Google Scholar Find this author on PubMed Search for this author on this site Issam ElNaqa 4 Department of Machine Learning, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Issam ElNaqa Steven Eschrich 3 Department of Biostatistics and Bioinformatics, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Steven Eschrich Hanlee P. Ji 5 Department of Medicine, Stanford University , CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Hanlee P. Ji Noemi Andor 2 Department of Integrated Mathematical Oncology, H Lee Moffitt Cancer Center and Research Institute , Tampa, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Noemi Andor For correspondence: noemi.andor{at}gmail.com Abstract Full Text Info/History Metrics Supplementary material Data/Code Preview PDF Abstract Molecular assays can resolve clonal structure, but they are expensive and typically sparse in time, whereas phenotypic observations such as imaging can be collected frequently but often are not preserved in the context needed for later interpretation. We present CLONEID , an event-based framework for organizing clone-resolved phenotypic, molecular, and specimen-context records so that genotype-to-phenotype interpretation can be maintained across time. CLONEID links time-stamped Events, assay-specific Perspectives, and reconciled Identities through structured ingestion, provenance-aware retrieval, and reproducible export, complementing upstream clone-calling methods. In a long-term gastric cancer density-selection experiment, CLONEID linked repeated culture events, growth measurements, and late karyotypic profiling within a shared record, supporting longitudinal interpretation of phenotypic adaptation together with underlying chromosomal state. Tumors and experimental cancer models often contain multiple co-existing clones whose phenotypic behavior and molecular composition can differ or change over time [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 ]. This intra-system heterogeneity creates a particularly informative setting for genotype-to-phenotype mapping, because related clones can differ in molecular state and behavior while still sharing much of their broader biological and environmental background. Here, we use clone operationally to mean a cell population inferred to share a common recent lineage and sufficiently similar underlying state to be treated as a single evolving unit across measurements, space, and time. Many methods infer clonal structure from sequencing and related assays [ 11 , 12 , 13 , 14 , 15 ]. By contrast, far fewer frameworks address the longitudinal data-integration problem: storing clone-resolved measurements together with the event and specimen context needed to connect phenotypic observations to later molecular readouts [ 16 , 17 ]. We developed CLONEID to address this problem as an event-based architecture and data model for capturing, linking, and reconciling longitudinal clone-resolved data across clinical, in vivo, and in vitro settings. CLONEID is organized around three core representational primitives—Event, Perspective and Identity—and one first-class measurement layer, Phenotype (Table ?? , Figure 1 ). Events record time-stamped propagation, sampling and intervention history together with specimen relationships and environmental context, thereby anchoring the history of a specimen in time and context. Perspectives are assay-specific molecular views of the specimen, typically partial and often obtained through destructive measurement, while preserving the provenance needed to interpret them. Because no single Perspective fully captures the machinery that gives rise to phenotype, CLONEID links multiple Perspectives to approximate an inferred clone-level Identity that can be followed across assays and over time. Phenotype refers to directly witnessed properties of the intact system. This includes phenotype state at a given event and behavioral phenotype inferred across repeated events over time. Molecular measurements, even when dynamic or rate-based, are not phenotype because they do not directly witness the intact system’s behavior or state; they provide assay-specific molecular perspectives on it. By linking phenotype directly to key events, CLONEID is designed to reduce a common failure mode in longitudinal studies: collecting molecular profiles without the contemporaneous phenotypic context needed for genotype-to-phenotype interpretation. Download figure Open in new tab Figure 1: Event-based architecture of CLONEID. (A) The same data model is used across clinical, in vivo and in vitro settings. Time-stamped Events capture specimen relationships, interventions and environmental context. (B) Longitudinal phenotype measurements anchor every event. (C) Assay-specific molecular readouts are represented as Perspectives , which preserve what was measured together with the upstream data required for provenance. (D) Multiple Perspectives are reconciled into an inferred Identity across assays and time. Together, this design enables integrated, queryable and reproducible retrieval of longitudinal genotype–phenotype records. Here we present the CLONEID architecture and its implementation for ingestion and retrieval of event-linked phenotypic and molecular records, with an emphasis on provenance-aware integration and reproducible access to longitudinal datasets. We highlight the system using a primary end-to-end demonstration in a long-term density-dependent selection experiment in gastric cancer cells, where repeated culture events and growth measurements informed subsequent karyotypic profiling and enabled longitudinal linkage of phenotypic adaptation to underlying chromosomal variations. An Event records “what happened, when, and to which specimen,” spanning propagation, sampling, and interventions across in vitro, in vivo, and clinical timelines ( Fig. 1A ). Each Event stores a time stamp, event type, specimen relationships, and contextual metadata, thereby anchoring downstream phenotypic and molecular records. CLONEID uses a common transition structure across study settings: genesis Events may have no parent, whereas non-genesis Events are linked to a precursor record so that entry into a new physical or contextual state remains connected to the history from which it arose. This rule ensures that Events record the state associated with the location being entered or exited, rather than merely the chronological order of records. This matters biologically because abrupt context changes, such as transplantation-like transitions, can alter the selective constraints acting on a clone and therefore require especially explicit contextual linkage. In practice, this particularly motivates Event creation when cells relocate into a new physical environment such as seeding, transfer, sampling, biopsy, resection, implantation, or intervention, but Event creation can also be justified by phenotype collection alone, even when no major biological transition has occurred. Because the same Event structure is used across clinical, in vivo and in vitro records, CLONEID represents these settings within a single scenario-neutral architecture ( Fig. 1A ). Phenotypes describe the state or behavior of cells or cell populations while the system remains intact and interpretable through its event history. In CLONEID, phenotype includes phenotype state , i.e. state at a given Event or timepoint, and behavioral phenotype , i.e. a temporal pattern that becomes interpretable only across repeated Events. Behavioral phenotype, by contrast, is inherently temporal and is inferred from repeated Event-linked evidence across time. Event-linked images, cell counts, and related non-molecular records provide observational evidence for phenotype state and also document that the Event occurred. Event-associated observational records, including imaging data, are converted into derived longitudinal measurements that describe phenotype such as segmentation-based features and related image-derived readouts within the same record structure. Phenotypes are by definition holistic, they capture how the intact system as a whole behaves or the state it is in. In contrast to Phenotypes, Perspectives are assay-specific and inherently partial readouts: each captures only one bounded view of the underlying biological machinery at a particular sampling moment. Perspectives ( Fig. 1C ) can store sample-level summaries, clone-level profiles and individual-cell measurements, while preserving the upstream data needed for provenance and reproducibility. CLONEID can retain both clone-resolved summaries and the underlying measurements from which they were derived, making each Perspective traceable, reproducible and available for downstream feature computation. Because no single Perspective exhausts what a cell is, multiple Perspectives need to be reconciled to infer Identity , the clone-level object that persists across contexts and gives rise to different phenotypes ( Fig. 1D ). This distinction is central to CLONEID’s larger goal of enabling genotype-to-phenotype mapping. Whereas a Perspective records what was measured in a specific assay, an Identity records the inferred clone-level object generated by linking one or more Perspectives that are considered consistent with the same underlying clone across assays, samples or time points. This reconciliation is encoded as an explicit relational layer: Perspectives remain assay-specific records, and Identity stores the links that group them into a shared clone representation while preserving their original provenance. These linked records are stored in an SQL relational database, accessible through two complementary interfaces: a web portal for interactive browsing, upload and export of longitudinal records, and an R package ( thecloneredesignlab/cloneidR ) for programmatic ingestion, Perspective registration, Identity reconciliation, querying and visualization. Both interfaces operate on the same event-based data model, allowing event-linked phenotypic records and assay-specific molecular Perspectives to be ingested, reconciled and retrieved within a common schema. To support portable and reproducible use, CLONEID preserves provenance at the level of both Events and Perspectives and provides context-bundled export of selected record subsets. In the current implementation, exported datasets retain the linkage between event history, phenotypic observations and downstream molecular records, allowing downstream analysis without requiring direct access to the full database. We used a long-term density-selection experiment in SNU-668 gastric cancer cells as a primary demonstration because it exercises all four pillars of CLONEID in a single setting: repeated Events, longitudinal phenotype capture, late molecular Perspectives and their reconciliation within a shared experimental history. Cells were maintained for more than six months under either low-confluence (r-selection) or high-confluence (K-selection) conditions, generating event-linked growth measurements and a late karyotypic readout for each adaptation history ( Fig. 2 ). Growth trajectories diverged by selection regime ( Fig. 2D ), and late karyotypic profiles distinguished adaptation histories, including induction and expansion of whole-genome doubled cells in an r-selection regime ( Fig. 2E,F ). Together, these data suggest that specific karyotypic states are selected under distinct density conditions and contribute to context-specific phenotypic adaptation. Additional appendix analyses show that the same record structure also supports distinct downstream study designs, including prediction of long-term adaptation and cellular plasticity from transcriptomic snapshots (Fig. S4) and mathematical inference of patient-specific karyotype fitness landscapes from multimodal longitudinal glioblastoma records (Fig. S5). Download figure Open in new tab Figure 2: CLONEID links event history, growth dynamics and karyotypic state in a longitudinal density-adaptation experiment. (A) Event history of the full (r/K) adaptation subtree rooted at the parental SNU-668 lineage is shown to place all downstream trajectories in their common ancestry context. Karyotyping timepoints are marked by ‘X’. The branch leading to the image-bearing subset in (B) is highlighted. (B) A circular view of a K-selected sublineage shows representative microscopy overlays together with the corresponding event history and passage progression. (C-D) Phenotypes . (C) Daily cell counts and fitted growth curves are shown for matched passages across replicates adapting to sparse (r-selection) or dense (K-selection) population conditions. (D) Growth rate and maximum population size are summarized across passages for all r- and K-selected replicates. (E,F) Perspective and Identity profiles distinguish adaptation histories in SNU-668. (E) Mean autosomal copy number per cell is shown as a genome-derived surrogate for ploidy across late control, r-selected, and K-selected specimens. (F) Cell-level karyotypic Perspectives resolve recurrent, condition-associated Identity classes across independent replicates. Code to recreate this figure from data available through dev.cloneid.org is available at https://github.com/thecloneredesignlab/cloneidR/tree/master/analysis . As of this publication, the CLONEID database contains more than 6,400 event-anchored specimen records organized into 17 rooted lineages across clinical, in vivo and in vitro histories. Data types range from bulk and single-cell DNA/RNA-derived molecular perspectives to image-derived phenotypes. More than 30,000 microscopy, digital pathology and medical images are stored with their associated event context and specimen relationships, along with molecular profiles from over 100,000 individual cells (Fig. S3A). We envision a future in which structured phenotypic data are shared alongside genomic and transcriptomic data as a routine part of cancer research, including manuscript submission and grant-supported data sharing. CLONEID is designed as a step toward that future by providing an event-based framework that links phenotypic measurements to specimen history, assay-specific molecular views and inferred clone identity with preserved provenance. By making these data easier to capture, organize and retrieve across clinical, in vivo and in vitro settings, CLONEID can help turn longitudinal phenotype–genotype records into reusable scientific assets for reproducibility, cross-study integration and future AI-enabled interpretation. View this table: View inline View popup Download powerpoint Table 1: Ontology-oriented and practical definitions of the core CLONEID concepts. The “ontology definition” column states what sort of thing each concept is in the data model. The “practical / user-facing definition” column states how a user should interpret or instantiate that concept in practice. Online Methods Software architecture and implementation CLONEID is implemented as a three-tier system consisting of (i) a user-facing web portal for visualization, upload and export, (ii) a core functionality layer (implemented in Java in the current deployment) for ingestion, linkage and reconciliation of records, and (iii) an SQL database that stores Event-linked phenotypic measurements together with assay-specific molecular Perspectives and their associated meta-data. Uploaded data are managed within an access-controlled warehousing model. Phenotypic image data can be processed using version-controlled analysis applications (for example, Cellpose [ 18 ]) so that derived measurements are reproducible and traceable to the originating uploaded observations. Demonstration dataset: long-term density-selection experiment To demonstrate CLONEID in an end-to-end setting, we used a long-term density-selection experiment in the SNU-668 gastric cancer cell line. Cells were cultured in RPMI-1640 supplemented with 10% fetal bovine serum and 1% streptomycin-penicillin at 37°C. Two density-selection regimes were maintained over more than six months: low-confluence ( r -selection) and high-confluence ( K -selection). Three replicates were established per regime. Under r -selection, replicates were seeded at 135 cells/cm 2 and transferred upon exceeding 4 × 10 3 cells/cm 2 , returning the population to 135 cells/cm 2 at each seeding. Under K -selection, replicates were seeded at 10 5 cells/cm 2 and transferred upon exceeding 2 × 10 5 cells/cm 2 . In CLONEID, each seeding and transfer was recorded as an Event linked to its source specimen and contextual metadata. Perspectives and Phenotypes CLONEID distinguishes Phenotype from Perspective by the type of information represented rather than by the measurement modality used to obtain it. Directly witnessed event-level state summaries belong to the Phenotype layer, whereas event-linked images of intact specimens are stored as observational evidence supporting phenotype interpretation. Measurements representing a molecular layer such as the genome, transcriptome, proteome, or karyotype are stored as Perspectives, including when those measurements arise from image-based assays. Spatial molecular assays therefore remain Perspectives, with coordinates stored as part of the molecular record rather than treated as purely phenotypic observations. Event-linked perspective capture The common Perspective representation is intentionally defined at the level of the molecular layer being profiled rather than the assay platform used to measure it. CLONEID does not require assay-specific ingestion routes because it does not ingest assays at the level of platform-specific file formats. Instead, it ingests a common event-linked Perspective representation: a biosample anchored to an Event, a set of inferred subpopulations with weights, and assay-specific feature matrices describing those subpopulations and optionally their members with spatial coordinates when relevant. This abstraction is broad enough to accommodate the shared core outputs of bulk clonal inference, single-cell clone assignment, and many spatial transcriptomic analyses after conversion into a common representation (Appendix 1.3), while method-specific auxiliary outputs such as trees and uncertainty may require additional metadata layers. Event-linked phenotype capture and growth measurements Phenotypic measurements were derived from event-linked observations recorded throughout the experiment. Bright-field images were uploaded and linked to their corresponding Events, and image-derived measurements, including segmentation-based features and cell counts were stored within the CLONEID record structure. Daily cell counts were used to generate fitted growth curves and growth-rate estimates across culture events. These longitudinal measurements were then queried directly from CLONEID for visualization and downstream comparison between adaptation histories. Derived phenotype quantities and processing assumptions Several phenotype-associated quantities stored in CLONEID are derived rather than directly recorded (Table 2). For microscopy-based culture events, stored cell-count fields are estimated from segmented image fields and then scaled to the full vessel surface area using the recorded flask area. These values therefore represent whole-vessel approximations rather than direct counts of all cells in the culture vessel. Likewise, stored area-occupied values are extrapolated from segmented image fields to the full vessel area, and stored cell-size values are summary statistics derived from segmented cell areas rather than direct per-cell measurements. In the current implementation, cell size reflects a high quantile of per-image segmented-cell areas aggregated across images, rather than a direct mean size measurement. Image-derived phenotype summaries may also be altered before final event-level values are stored. In the current pipeline, microscopy images are preprocessed before segmentation by default, including cell-line-aware gamma correction unless preprocessing is explicitly disabled. In addition, image-derived cell-count summaries may be filtered by an image-quality model and, when flagged, by manual exclusion of selected image positions before flask-level values are finalized. Microscopy segmentation parameters are also not fixed globally: cell-line-specific and, in some cases, lineage-specific parameter sets can be selected automatically, with a default parameter set used when no lineage-specific match is found. Event-history curation and modality-specific caveats For workflows in which event identifiers encode ordered transitions, CLONEID uses two partially redundant sources of lineage information: the explicit parent–child linkage stored in the database and the implicit transition structure encoded in user-defined event identifiers. This redundancy is used as a validation mechanism rather than as an independent source of truth. In the current implementation, when an event identifier contains a single passage-designating token of the form A (for example A1 , A7 , or A24 ), the insertion logic combines that token with event type and event date to infer the predecessor that would be expected under the transition sequence. More generally, the logic distinguishes arrival-events , which introduce a specimen into a new physical or contextual state, from departure-events , which precede or terminate such a transition. For arrival-events, the expected parent is a departure-event from the preceding stage; for departure-events, the expected parent is the corresponding arrival-event within the same stage that occurred earlier in time. Insertion is aborted only when this inferred parent is unique and conflicts with the supplied parent link. If the identifier does not support a unique inference, or if multiple candidate parents remain plausible, the system emits only a warning or message and does not automatically reject the record. This design aims to catch likely lineage-entry mistakes in long-term studies while avoiding overly rigid constraints on valid but ambiguously named records. Some event metadata may also be inherited by default: for example, departure events inherit flask identity from the parent event, and media information may be copied from the parent event unless specified explicitly. Some non-microscopy modalities reuse phenotype storage fields with modality-specific meanings. In the current MRI path, a mask-derived volumetric quantity is stored through the same workflow that otherwise records microscopy-derived size summaries. These fields should therefore be interpreted in a modality-aware manner rather than assumed to represent the same physical quantity across all event types. Late karyotypic profiling as a molecular Perspective At late stages of adaptation, samples from each of the six test trajectories were collected for karyotyping together with two parental controls. Karyotypic profiles were obtained for 28 K -selected cells, 61 r -selected cells and 32 control cells, for a total of 121 cells. Samples were collected after 27 serial culture events under K -selection and 55 serial culture events under r -selection. The cumulative area of segmented chromosomes was used as a surrogate for ploidy, and chromosome-level copy-number profiles were used for clustering and comparison of adaptation histories. In CLONEID, these karyotype-derived readouts were represented as late molecular Perspectives linked back to their event history and longitudinal phenotypic measurements. Preprocessing of genomic and karyotype data scRNA-seq karyotype inference For patient-derived analyses in the Appendix, we inferred changes in karyotype composition from single-cell transcriptomic data using Numbat, enabling comparison of inferred CNA states between primary and recurrent specimens. Karyotype image processing For karyotype images used in the density-selection demonstration and in Appendix analyses, karyotype panels were processed to segment individual chromosomes and assign chromosome labels. Karyotype images were required to contain chromosome groups arranged by standard karyotyping conventions with chromosome numbers annotated beneath each group. The analysis pipeline (i) detected annotated chromosome numbers and their image locations using optical character recognition, (ii) segmented chromosomes by thresholding and assigned each segmented chromosome to a chromosome group based on the detected label locations, and (iii) measured chromosome-level features including area (used as a ploidy surrogate in this work) and arm-level properties using morphological processing and skeletonization-derived endpoints. Mathematical modeling For the Appendix patient cohort analysis, we used ALFA-K [ 19 ] as an external downstream modeling framework to infer local fitness landscapes for aneuploid karyotypes from single-cell karyotype data. CLONEID was used to store and organize the event-linked clinical and molecular records used as inputs to modeling (for example, imaging time points, biopsy Events and single-cell derived karyotype states), while fitness landscape inference and cross-validation were performed by ALFA-K. Interpretation of the demonstration analysis Clustering of late karyotypic profiles separated cells according to their selection history. Most test-condition cells regained a copy of chromosome 21 relative to parental controls, and additional differences between r - and K -selected cells included loss of chromosomes 10 and 14 in the K -selected group. Karyotyping also identified tetraploidized cells enriched under r -selection, including whole-genome doubling in the r2 trajectory. Within CLONEID, these data were represented as late molecular Perspectives linked back to the corresponding event history and longitudinal phenotypic measurements. Funding This work was supported by the NCI grants 1R03CA259873-01A1, 1R37CA266727-01A1 and 1R21CA269415-01A1 awarded to N. Andor. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Code availability CLONEID is accessible through the web portal at dev.cloneid.org, which provides interactive access to the current CLONEID resource, including linked phenotypic, molecular, and specimen-context records. The portal also supports export of context-preserving subsets for downstream analysis. For users who wish to implement the CLONEID storage model within their own in-house database deployment, the corresponding cloneidR package is also available at https://github.com/thecloneredesignlab/cloneidR.git . Data availability The data currently represented in CLONEID, including linked phenotypic, molecular, and specimen-context records, are accessible through the CLONEID portal at dev.cloneid.org. The portal supports interactive querying and export of context-linked subsets so that event history, specimen relationships, and associated molecular and phenotypic data remain connected during downstream reuse. Acknowledgments We thank Sue Grimes for her early contributions to the database design, which helped ensure its long-term scalability. Funder Information Declared NCI , 1R21CA269415-01A1 , 1R37CA266727-01A1 , 1R03CA259873-01A1 Footnotes This version has been revised to sharpen the central motivation and improve clarity, accessibility, and practical detail. The title and abstract now frame CLONEID more directly as a longitudinal data framework for context specific genotype to phenotype mapping. The introduction has been rewritten to make the main methodological need clearer, namely preserving the event and specimen context required to connect phenotypic observations to later molecular readouts. The conceptual model has been clarified throughout. Event is now presented as the primary public facing organizing unit, with more explicit definitions of Event, Perspective, Identity, Phenotype, Clone, and Specimen or study entity. The distinctions between phenotype, perspective, and identity have been substantially revised to make clear that phenotypes capture behavior of the intact system in context, perspectives are assay specific partial molecular views, and identities reconcile one or more perspectives into an inferred clone level representation. The manuscript also now places clinical, in vivo, and in vitro settings on more equal footing. Event definitions, examples, and workflow descriptions have been expanded to show how the same event based architecture applies across these domains. Figure 1 and its accompanying text have been revised to better convey the Event, Perspective, Identity framework. The primary use case figure and text have also been updated to make clearer how late karyotypic profiles support an inferred identity like interpretation in the density selection experiment. Additional revisions strengthen the practical software and resource description. The manuscript now includes expanded material on user facing and programmatic interfaces, ingestion workflow, file conventions for molecular perspectives, context bundled export, access modes, storage model, and current database content. The storage description has been updated to reflect the current separation of relational metadata from larger phenotype artifacts, including S3 backed storage for imaging and segmentation outputs. Quantitative resource summary statements have been refreshed. Appendix and methods material have been expanded to provide more explicit implementation details, caveats, and examples of derived quantities. Overall, the revised version places greater emphasis on practical value, longitudinal context preservation, and external usability, while retaining the original biological demonstrations. https://dev.cloneid.org References [1]. ↵ M. Greaves and C. C. Maley . Clonal evolution in cancer . Nature , 481 ( 7381 ): 306 – 313 , January 2012 . OpenUrl CrossRef PubMed Web of Science [2]. ↵ J. P. O’Connor , C. J. Rose , J. C. Waterton , R. A. Carano , G. J. Parker , and A. Jackson . Imaging Intratumor Heterogeneity: Role in Therapy Response, Resistance, and Clinical Outcome . Clinical cancer research : an official journal of the American Association for Cancer Research , 21 ( 2 ): 249 – 257 , January 2015 . PMID: 25421725 . OpenUrl Abstract / FREE Full Text [3]. ↵ A. Kreso , C. A. O’Brien , P. van Galen , O. I. Gan , F. Notta , A. M. K. Brown , K. Ng , J. Ma , E. Wienholds , C. Dunant , A. Pollett , S. Gallinger , J. McPherson , C. G. Mullighan , D. Shibata , and J. E. Dick . Variable clonal repopulation dynamics influence chemotherapy response in colorectal cancer . Science (New York, N.Y .), 339 ( 6119 ): 543 – 548 , February 2013 . PMID: 23239622 . OpenUrl Abstract / FREE Full Text [4]. ↵ U. Ben-David , B. Siranosian , G. Ha , H. Tang , Y. Oren , K. Hinohara , C. A. Strathdee , J. Dempster , N. J. Lyons , R. Burns , A. Nag , G. Kugener , B. Cimini , P. Tsvetkov , Y. E. Maruvka , R. O’Rourke , A. Garrity , A. A. Tubelli , P. Bandopadhayay , A. Tsherniak , F. Vazquez , B. Wong , C. Birger , M. Ghandi , A. R. Thorner , J. A. Bittker , M. Meyerson , G. Getz , R. Beroukhim , and T. R. Golub . Genetic and transcriptional evolution alters cancer cell line drug response . Nature , 560 ( 7718 ): 325 – 330 , August 2018 . PMID: 30089904 . OpenUrl CrossRef PubMed [5]. ↵ U. Ben-David , G. Ha , Y.-Y. Tseng , N. F. Greenwald , C. Oh , J. Shih , J. M. Mc-Farland , B. Wong , J. S. Boehm , R. Beroukhim , and T. R. Golub . Patient-derived xenografts undergo mouse-specific tumor evolution . Nature Genetics , 49 ( 11 ): 1567 – 1575 , November 2017 . PMID: 28991255 . OpenUrl CrossRef PubMed [6]. ↵ T. Li , J. Liu , J. Feng , Z. Liu , S. Liu , M. Zhang , Y. Zhang , Y. Hou , D. Wu , C. Li , Y. Chen , H. Chen , and X. Lu . Variation in the life history strategy underlies functional diversity of tumors . National Science Review , 8 ( waa124 ), February 2021 . [7]. ↵ P. Hughes , D. Marshall , Y. Reid , H. Parkes , and C. Gelber . The costs of using unauthenticated, over-passaged cell lines: how much more data do we need? BioTechniques , 43 ( 5 ): 575 , 577–578, 581–582 passim, November 2007 . PMID: 18072586 . OpenUrl CrossRef PubMed [8]. ↵ N. Andor , B. T. Lau , C. Catalanotti , A. Sathe , M. A. Kubit , J. Chen , S. M. Grimes , C. J. Suarez , and H. P. Ji . Joint single cell DNA-seq and RNA-seq of gastric cancer cell lines reveals rules of in vitro evolution . NAR Genomics and Bioinformatics (in press) , March 2020 . [9]. ↵ Y. Liu , Y. Mi , T. Mueller , S. Kreibich , E. G. Williams , A. Van Drogen , C. Borel , M. Frank , P.-L. Germain , I. Bludau , M. Mehnert , M. Seifert , M. Emmenlauer , Sorg, F. Bezrukov , F. S. Bena , H. Zhou , C. Dehio , G. Testa , J. Saez-Rodriguez , S. E. Antonarakis , W.-D. Hardt , and R. Aebersold . Multi-omic measurements of heterogeneity in HeLa cells across laboratories . Nature Biotechnology , 37 ( 3 ): 314 – 322 , March 2019 . OpenUrl CrossRef PubMed [10]. ↵ D. C. Minussi , M. D. Nicholson , H. Ye , A. Davis , K. Wang , T. Baker , M. Tarabichi , E. Sei , H. Du , M. Rabbani , C. Peng , M. Hu , S. Bai , Y.-W. Lin , A. Schalck , A. Multani , J. Ma , T. O. McDonald , A. Casasent , A. Barrera , H. Chen , B. Lim , B. Arun , F. Meric-Bernstam , P. Van Loo , F. Michor , and N. E. Navin . Breast tumours maintain a reservoir of subclonal diversity during expansion . Nature , 592 ( 7853 ): 302 – 308 , April 2021 . PMID: 33762732 . OpenUrl CrossRef PubMed [11]. ↵ C. Ma , M. Balaban , J. Liu , S. Chen , M. J. Wilson , C. H. Sun , L. Ding , and B. J. Raphael . Inferring allele-specific copy number aberrations and tumor phylogeography from spatially resolved transcriptomics . Nature Methods , 21 ( 12 ): 2239 – 2247 , December 2024 . OpenUrl PubMed [12]. ↵ S. Zaccaria and B. J. Raphael . Accurate quantification of copy-number aberrations and whole-genome duplications in multi-sample tumor sequencing data . Nature Communications , 11 ( 1 ): 4301 , September 2020 . OpenUrl PubMed [13]. ↵ A. Roth , J. Khattra , D. Yap , A. Wan , E. Laks , J. Biele , G. Ha , S. Aparicio , Bouchard-Côté, and S. P. Shah . PyClone: statistical inference of clonal population structure in cancer . Nature Methods , 11 ( 4 ): 396 – 398 , April 2014 . OpenUrl PubMed [14]. ↵ P. Patel , I. Tirosh , J. J. Trombetta , A. K. Shalek , S. M. Gillespie , H. Wakimoto , D. P. Cahill , B. V. Nahed , W. T. Curry , R. L. Martuza , D. N. Louis , O. Rozenblatt-Rosen , M.L. Suvá , A. Regev , and B. E. Bernstein . Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma . Science (New York, N.Y .), 344 ( 6190 ): 1396 – 1401 , June 2014 . PMID: 24925914 . OpenUrl Abstract / FREE Full Text [15]. ↵ N. Navin , J. Kendall , J. Troge , P. Andrews , L. Rodgers , J. McIndoo , K. Cook , A. Stepansky , D. Levy , D. Esposito , L. Muthuswamy , A. Krasnitz , W. R. McCombie , Hicks, and M. Wigler . Tumour evolution inferred by single-cell sequencing . Nature , 472 ( 7341 ): 90 – 94 , April 2011 . OpenUrl CrossRef PubMed Web of Science [16]. ↵ S. H. Selznick , M. L. Thatcher , K. S. Brown , and C. A. Haussler . Development and application of computer software for cell culture laboratory management . In Vitro Cellular & Developmental Biology. Animal , 37 ( 1 ): 55 – 61 , January 2001 . PMID: 11249207 . OpenUrl PubMed [17]. ↵ S. G. Canfield , G. Jin , S. P. Palecek , and T. Sampsell . Software to improve transfer and reproducibility of cell culture methods . BioTechniques , 65 ( 5 ): 289 – 292 , November 2018 . OpenUrl CrossRef PubMed [18]. ↵ Stringer , T. Wang , M. Michaelos , and M. Pachitariu . Cellpose: a generalist algorithm for cellular segmentation . Nature Methods , 18 ( 1 ): 100 – 106 , January 2021 . OpenUrl CrossRef PubMed [19]. ↵ R. J. Beck and N. Andor . Local Adaptive Mapping of Karyotype Fitness Landscapes , July 2023 . Pages: 2023.07.14.549079 Section: New Results. [20]. H. S. Kim , S. M. Grimes , T. Chen , A. Sathe , B. T. Lau , G.-H. Hwang , S. Bae , and H. P. Ji . Direct measurement of engineered cancer mutations and their transcriptional phenotypes in single cells . Nature Biotechnology , 42 ( 8 ): 1254 – 1262 , August 2024 . OpenUrl CrossRef PubMed [21]. S. Müller , A. Cho , S. J. Liu , D. A. Lim , and A. Diaz . CONICS integrates scRNA-seq with DNA sequencing to map gene expression to tumor sub-clones . Bioinformatics (Oxford, England) , 34 ( 18 ): 3217 – 3219 , September 2018 . PMID: 29897414 . OpenUrl CrossRef PubMed [22]. S. Müller , E. D. Lullo , A. Bhaduri , B. Alvarado , G. Yagnik , G. Kohanbash , M. Aghi , and A. Diaz . A single-cell atlas of human glioblastoma reveals a single axis of phenotype in tumor-propagating cells . bioRxiv , page 377606 , July 2018 . [23]. Wang , J. Jung , H. Babikir , K. Shamardani , S. Jain , X. Feng , N. Gupta , S. Rosi , S. Chang , D. Raleigh , D. Solomon , J. J. Phillips , and A. A. Diaz . A single-cell atlas of glioblastoma evolution under therapy reveals cell-intrinsic and cell-extrinsic therapeutic targets . Nature Cancer , 3 ( 12 ): 1534 – 1552 , December 2022 . OpenUrl PubMed View the discussion thread. Back to top Previous Next Posted April 08, 2026. Download PDF Supplementary Material Data/Code Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following A longitudinal data framework for context-specific genotype-to-phenotype mapping Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share A longitudinal data framework for context-specific genotype-to-phenotype mapping Thomas Veith , Richard Beck , Vural Tagal , Tao Li , Saeed Alahmari , Jackson Cole , Daniel Hannaby , John Kyei , Xiaoqing Yu , Konstantin Maksin , Andrew Schultz , HoJoon Lee , Aaron Diaz , Janine Lupo , Issam ElNaqa , Steven Eschrich , Hanlee P. Ji , Noemi Andor bioRxiv 2025.05.07.652202; doi: https://doi.org/10.1101/2025.05.07.652202 Share This Article: Copy Citation Tools A longitudinal data framework for context-specific genotype-to-phenotype mapping Thomas Veith , Richard Beck , Vural Tagal , Tao Li , Saeed Alahmari , Jackson Cole , Daniel Hannaby , John Kyei , Xiaoqing Yu , Konstantin Maksin , Andrew Schultz , HoJoon Lee , Aaron Diaz , Janine Lupo , Issam ElNaqa , Steven Eschrich , Hanlee P. Ji , Noemi Andor bioRxiv 2025.05.07.652202; doi: https://doi.org/10.1101/2025.05.07.652202 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Bioinformatics Subject Areas All Articles Animal Behavior and Cognition (7618) Biochemistry (17636) Bioengineering (13860) Bioinformatics (41847) Biophysics (21401) Cancer Biology (18536) Cell Biology (25424) Clinical Trials (138) Developmental Biology (13353) Ecology (19860) Epidemiology (2067) Evolutionary Biology (24287) Genetics (15583) Genomics (22463) Immunology (17701) Microbiology (40300) Molecular Biology (17141) Neuroscience (88434) Paleontology (666) Pathology (2825) Pharmacology and Toxicology (4813) Physiology (7633) Plant Biology (15107) Scientific Communication and Education (2042) Synthetic Biology (4285) Systems Biology (9808) Zoology (2268)