CM4AI Dataset Documentation

Description	ID	Name
Deliver machine-readable hierarchical maps of cell architecture as AI-Ready data from multimodal int...	purpose-001	AI-Ready Cell Architecture Maps for Biomedical AI
Address the grand challenge of interpretable genotype-phenotype learning in genomics and precision m...	purpose-002	Interpretable Genotype-Phenotype Learning
Establish standards, best practices, and guidelines for ethical AI-readiness in biomedical data. Thi...	purpose-003	Establishing AI-Readiness Standards for Biomedical Data

📊

Composition

What do the instances represent?

Instances

Description	ID	Instance Type	Name
Triple negative breast cancer cell line (RRID:CVCL_0419) established from a metastatic site pleural ...	instance-001	Cultured cell line from de-identified human tissue, ethically sourced from ATCC	MDA-MB-468 Breast Cancer Cell Line
Human induced pluripotent stem cell (iPSC) line (RRID:CVCL_B5P3) derived from a healthy male Norther...	instance-002	Cultured cell line from de-identified human tissue, ethically sourced from HipSci	KOLF2.1J Induced Pluripotent Stem Cells
Near-comprehensive set of chromatin regulators encoded by the human genome analyzed via AP-MS, SEC-M...	instance-003	Protein targets for multi-modal analysis	100 Chromatin Regulators
Set of metabolic enzymes involved in cancer, neuropsychiatric, and cardiac disorders analyzed via mu...	instance-004	Protein targets for multi-modal analysis	100 Metabolic Enzymes

Subpopulations

Description	ID	Name
MDA-MB-468 breast cancer cells in control/untreated condition	subpop-001	MDA-MB-468 Untreated
MDA-MB-468 breast cancer cells treated with paclitaxel chemotherapy	subpop-002	MDA-MB-468 Paclitaxel-Treated
MDA-MB-468 breast cancer cells treated with vorinostat chemotherapy	subpop-003	MDA-MB-468 Vorinostat-Treated
KOLF2.1J induced pluripotent stem cells in undifferentiated/naive state	subpop-004	KOLF2.1J Undifferentiated iPSCs
KOLF2.1J iPSCs differentiated into neurons	subpop-005	KOLF2.1J iPSC-Derived Neurons
KOLF2.1J iPSCs differentiated into neural progenitor cells	subpop-006	KOLF2.1J iPSC-Derived Neural Progenitor Cells (NPCs)
KOLF2.1J iPSCs differentiated into cardiomyocytes	subpop-007	KOLF2.1J iPSC-Derived Cardiomyocytes

Distribution Formats

Access Urls	Description	ID	Name
https://doi.org/10.18130/V3/DXWOS5, https://doi.org/10.18130/V3/B35XWX, ... (+2 more)	All CM4AI output data packaged as Research Object Crate (RO-Crate) packages containing datasets, met...	format-001	RO-Crate Packages with Provenance
MassIVE Repository (human iPSCs), MassIVE Repository (human cancer cells)	Mass spectrometry data deposited to MassIVE Repository (Proteomics community-supported repository). ...	format-002	Mass Spectrometry Data in MassIVE
NCBI BioProject, Sequence Read Archive (SRA)	Raw sequence data from CRISPR perturbation screens deposited to NCBI BioProject/Sequence Read Archiv...	format-003	Sequence Data in NCBI SRA
https://www.ndexbio.org	Cell maps shared via Network Data Exchange (NDEx) for visualization and access. Maps can be visualiz...	format-004	Hierarchical Cell Maps in NDEx
https://doi.org/10.18130/V3/DXWOS5, LibraData University of Virginia	Archived RO-Crates available in University of Virginia's LibraData data archive (instance of Harvard...	format-005	University of Virginia Dataverse

🔍

Collection Process

How was the data acquired?

ID *

https://doi.org/10.18130/V3/DXWOS5

Name

CM4AI

Title

Cell Maps for Artificial Intelligence (CM4AI)

Description

CM4AI is the Functional Genomics Data Generation Project in the U.S. National Institutes of Health's (NIH) Bridge to Artificial Intelligence (Bridge2AI) program. Its overarching mission is to produce ethical, AI-ready datasets of cell architecture, inferred from multimodal data collected for human cell lines, to enable transformative biomedical AI research. The project delivers machine-readable hierarchical maps of cell architecture as AI-Ready data produced from multimodal interrogation of 100 chromatin modifiers and 100 metabolic enzymes involved in cancer, neuropsychiatric, and cardiac disorders in disease-relevant cell lines under perturbed and unperturbed conditions. Data streams include immunofluorescence (IF) subcellular microscopy for spatial proteomics, affinity purification mass spectroscopy (AP-MS) and size exclusion mass spectroscopy (SEC-MS) for protein-protein interaction (PPI) data, and single-cell CRISPR-Cas perturbation screens by cell type. Input data streams are integrated via the Multi-Scale Integrated Cell (MuSIC) software pipeline employing deep learning models and community detection algorithms, and output cell maps are packaged with provenance graphs and rich metadata as AI-Ready datasets in RO-Crate format using the FAIRSCAPE framework.

Page

https://www.cm4ai.org

Language

Keywords

Cell Maps
Artificial Intelligence
AI-Ready Data
Bridge2AI
Functional Genomics
Protein-Protein Interactions
Spatial Proteomics
CRISPR Perturbation
Hierarchical Cell Maps
MDA-MB-468
KOLF2.1J
iPSC
Breast Cancer
Chromatin Modifiers
Metabolic Enzymes
Immunofluorescence
Mass Spectrometry
AP-MS
SEC-MS
Perturb-Seq
FAIR Principles
RO-Crate
FAIRSCAPE
Visible Neural Networks
Deep Learning

Addressing Gaps

Description	ID	Name
Address the limitation that machine learning models in genomics and precision medicine are typically...	gap-001	Black Box AI Models in Genomic Medicine
Provide fully provenanced, ethically validated, and FAIR-compliant AI-ready datasets with machine-re...	gap-002	Lack of AI-Ready Biomedical Datasets with Provenance
Create integrated datasets combining protein localization (spatial proteomics), protein-protein inte...	gap-003	Integration of Multimodal Cellular Data

Creators

Role	Name	ORCID	Affiliation
Contributor	Trey Ideker	creator-001	-
Contributor	Jean-Christophe Bélisle-Pipon	creator-002	-
Contributor	Timothy Clark	creator-003	-
Contributor	Jake Yue Chen	creator-004	-
Contributor	Nevan J Krogan	creator-005	-
Contributor	Emma Lundberg	creator-006	-
Contributor	Prashant Mali	creator-007	-
Contributor	Sarah J Ratcliffe	creator-008	-
Contributor	Vardit Ravitsky	creator-009	-
Contributor	Andrej Sali	creator-010	-
Contributor	Wade Loren Schulz	creator-011	-
Contributor	Ying Ding	creator-012	-
Contributor	Samah Fodeh	creator-013	-
Contributor	Cynthia Brandt	creator-014	-
Contributor	Pamela Payne-Foster	creator-015	-

Subsets

Description	ID	Name
Immunofluorescence-based staining (ICC-IF) and confocal microscopy images displaying spatial localiz...	subset-001	Spatial Proteomics IF Images - MDA-MB-468
Affinity purification mass spectrometry (AP-MS) data on endogenously tagged cell lines mapping prote...	subset-002	Protein-Protein Interaction AP-MS Data
Size exclusion chromatography coupled to mass spectrometry (SEC-MS) for proteome-wide complex/ inter...	subset-003	Protein-Protein Interaction SEC-MS Data
Genome-scale CRISPRi perturbation cell atlas in undifferentiated KOLF2.1J human induced pluripotent ...	subset-004	CRISPR Perturbation Cell Atlas
Integrated hierarchical cell maps produced by Multi-Scale Integrated Cell (MuSIC) pipeline from fusi...	subset-005	Hierarchical Cell Maps via MuSIC

Sampling Strategies

ID
sampling-001
Name
Disease-Relevant Cell Line Selection
Description
Purposive selection of two disease-relevant cell lines: MDA-MB-468 triple negative breast cancer cell line for cancer research, and KOLF2.1J iPSCs for neuropsychiatric and cardiac disorder research. Both cell lines ethically sourced and well-characterized in the literature.
Is Sample
- False
Is Random
- False
Is Representative
- False
Strategies
- Selection of commercially available, ethically sourced cell lines
- MDA-MB-468 chosen for triple-negative breast cancer research applications
- KOLF2.1J chosen as reference iPSC line for large-scale collaborative studies
- Both cell lines have extensive existing characterization data

Collection Mechanisms

Description	ID	Name
Automated fixation and permeabilization protocols using pipetting robot for MDA-MB-468 and KOLF2.1J ...	collection-001	Immunofluorescence Spatial Proteomics Imaging
Endogenous tagging of genes in cell lines followed by affinity purification mass spectrometry (AP-MS...	collection-002	Affinity Purification Mass Spectrometry
Size exclusion chromatography coupled to mass spectrometry (SEC-MS) for proteome-wide complex/ inter...	collection-003	Size Exclusion Chromatography Mass Spectrometry
Single-cell CRISPR screens using CRISPR lentiviral library targeting 100 chromatin factors with 6 gu...	collection-004	CRISPR Perturbation Screens

Acquisition Methods

Description	ID	Name
High-resolution confocal microscopy of immunofluorescence-stained cells capturing four channels: DAP...	acquisition-001	Confocal Microscopy for Subcellular Imaging
State-of-the-art mass spectrometry-based proteomics including AP-MS on endogenously tagged cell line...	acquisition-002	Mass Spectrometry for Protein Interactions
Single-cell RNA sequencing using 10x Genomics 3'HT kit to capture transcriptional states following C...	acquisition-003	Single-Cell RNA Sequencing for Perturbation Mapping
Multi-Scale Integrated Cell (MuSIC) pipeline integrates PPI embeddings and image embeddings using co...	acquisition-004	MuSIC Pipeline Integration

Preprocessing Strategies

Description	ID	Name	Preprocessing Details
PPI networks processed using node2vec deep learning model to reduce dimensionality and produce PPI e...	preproc-001	Deep Learning Embedding Generation	node2vec deep learning for PPI network dimensionality reduction, Human Protein Atlas deep learning model for image embedding, ... (+2 more)
Community detection performed on co-embedding space using multiscale community detection algorithms ...	preproc-002	Hierarchical Community Detection	All-by-all similarity computation in co-embedding space, Multiscale community detection using Cytoscape algorithms, ... (+2 more)
Two-pronged annotation approach: (1) Alignment to known protein function and pathway resources inclu...	preproc-003	Cell Map Annotation	Alignment to Gene Ontology for functional annotation, Alignment to Reactome pathways for pathway annotation, ... (+2 more)
Standardized automated fixation and permeabilization protocols using pipetting robot. Consistent sta...	preproc-004	Quality Control for Imaging Data	Automated protocols for consistency, Standardized staining across all conditions, ... (+2 more)
Quality control and validation of AP-MS and SEC-MS data. Data currently undergoing quality checks be...	preproc-005	Quality Control for Mass Spectrometry Data	QC procedures for AP-MS data, QC procedures for SEC-MS data, ... (+2 more)

Cleaning Strategies

Cleaning Details	Description	ID	Name
RO-Crate packaging with metadata and provenance, JSON-Schema validation of all datasets, ... (+3 more)	All datasets packaged using FAIRSCAPE framework which creates RO-Crate packages with datasets, metad...	cleaning-001	FAIRSCAPE AI-Readiness Packaging
GO mapping for protein functions, Reactome mapping for pathways, ... (+3 more)	Data mapped to applicable standards and ontologies including Gene Ontology (GO), Reactome, Protein D...	cleaning-002	Data Standard Mapping
PDB structural information integration, AlphaFoldDB structure integration, ... (+3 more)	Bioinformatics pipeline developed for annotating MuSIC communities with available structural informa...	cleaning-003	Integrative Structure Modeling Annotation

Maintainers

ID
maintainer-001
Name
CM4AI Consortium
Description
Multidisciplinary consortium managing dataset maintenance including University of California San Diego (lead), University of California San Francisco, Stanford University, University of Virginia, Yale University, University of Alabama at Birmingham, Simon Fraser University, and The Hastings Center. Data Governance Committee led by Jillian Parker. Ethical Review by Vardit Ravitsky and Jean-Christophe Belisle-Pipon.
Maintainer Details
- University of California San Diego (lead institution, Ideker Lab)
- UCSF (Krogan Lab - protein interactions, Sali Lab - structure modeling)
- Stanford University (Lundberg Lab - spatial proteomics)
- University of Virginia (Clark Lab - standards and FAIRSCAPE)
- Yale University (Schulz Lab - workforce development)
- University of Alabama at Birmingham (Chen Lab - teaming, U-BRITE platform)
- Simon Fraser University (Bélisle-Pipon - ethics)
- The Hastings Center (Ravitsky - ethics)
- Data Governance Committee (Jillian Parker)

Retention Limit

ID

retention-001

Name

Long-Term Preservation Plan

Description

Digital data maintained according to NIH data sharing policies with long-term preservation in University of Virginia's LibraData repository supported by committed institutional funds. No planned sunset for data availability. Archived RO-Crates with persistent identifiers (ARK, future DOIs) ensure long-term accessibility and citability.

Retention Details

NIH data sharing policy compliance
UVA Dataverse institutional commitment
Persistent identifiers (ARK, future DOIs)
No planned data sunset
Machine-readable metadata for long-term discoverability

Sensitive Elements

ID
sensitive-001
Name
Cell Line Origin Metadata
Description
While cell lines are de-identified and cannot be matched to specific individuals, metadata about cell line origins (age, sex, race of original donor) is retained for scientific context. This metadata does not constitute identifiable human subjects data under current knowledge.
Sensitive Elements Present
False
Sensitivity Details
- De-identified commercial cell lines
- Donor demographic metadata for scientific context only
- Cannot be matched to individuals with current knowledge
- Ethically sourced from ATCC and HipSci

External Resources

Description	External Resources	ID	Name
Official project website and data portal using U-BRITE platform	https://www.cm4ai.org	resource-001	CM4AI Project Website
Federal grant information and project details for Bridge2AI Functional Genomics	https://reporter.nih.gov/project-details/11211616	resource-002	NIH RePORTER Project Details
LibraData repository with archived RO-Crates and data releases	https://doi.org/10.18130/V3/DXWOS5, https://doi.org/10.18130/V3/B35XWX, https://doi.org/10.18130/V3/F3TD5R	resource-003	University of Virginia Dataverse
Clark T, et al. Cell Maps for Artificial Intelligence: AI-Ready Maps of Human Cell Architecture from...	https://doi.org/10.1101/2024.05.21.589311	resource-004	bioRxiv Publication
AI-readiness framework documentation, tutorial, and installation instructions	https://fairscape.github.io	resource-005	FAIRSCAPE Framework Documentation
Open source IMP package for integrative structure modeling	http://integrativemodeling.org	resource-006	Integrative Modeling Platform
Repository and visualization platform for cell maps and networks	https://www.ndexbio.org	resource-007	Network Data Exchange (NDEx)
Parent NIH Common Fund program supporting AI-ready biomedical datasets	https://commonfund.nih.gov/bridge2ai	resource-008	Bridge2AI Program
Collaboration partner for data curation and integration	https://www.nih-cfde.org	resource-009	NIH Common Fund Data Ecosystem (CFDE)
Mass spectrometry data repository for iPSC and cancer cell data	MassIVE Repository	resource-010	MassIVE Proteomics Repository
Repository for CRISPR perturbation screen raw sequence data	NCBI BioProject, Sequence Read Archive (SRA)	resource-011	NCBI Sequence Read Archive
Nourreddine S, et al. A PERTURBATION CELL ATLAS OF HUMAN INDUCED PLURIPOTENT STEM CELLS. bioRxiv. 20...	https://doi.org/10.1101/2024.11.03.621734	resource-012	Perturbation Cell Atlas Publication

🚀

Uses

What (other) tasks could the dataset be used for?

Tasks

Description	ID	Name
Integrate multimodal data streams (spatial proteomics via IF imaging, protein-protein interactions v...	task-001	Multi-Scale Cell Mapping via MuSIC Pipeline
Enable development of visible neural networks (VNNs) and visible machine learning tools that use hie...	task-002	Visible Neural Network Development
Characterize cell architecture and protein interactions in disease-relevant cell lines including tre...	task-003	Disease-Relevant Cell Line Characterization
Develop ethical AI frameworks and governance structures for biomedical data, including Value-Sensiti...	task-004	Ethical AI Framework Development

Intended Uses

Description	ID	Name
Primary intended use is training and development of artificial intelligence and machine learning mod...	use-001	AI Model Training for Functional Genomics
Development of visible neural networks (VNNs) that use hierarchical cell maps as interpretable model...	use-002	Visible Neural Network Development
Research into interpretable genotype-phenotype learning using multi-scale cell maps. Enables underst...	use-003	Genotype-Phenotype Mapping Research
Analysis of cellular responses to drug treatments (paclitaxel, vorinostat) to predict drug response ...	use-004	Drug Response and Synergy Prediction
Study of disease mechanisms in cancer, neuropsychiatric disorders, and cardiac disorders through ana...	use-005	Disease Mechanism Research
Use as exemplar for future AI-ready biomedical dataset development, demonstrating best practices in ...	use-006	Model for AI-Ready Biomedical Dataset Development

Discouraged Uses

Description	ID	Name
Laboratory data from cell lines are not to be used in clinical decision-making or any context involv...	discouraged-001	Clinical Decision-Making Without Validation
This is an interim/beta release with data not yet in completed final form. Some datasets are under t...	discouraged-002	Use During Incomplete Data Release
Datasets require domain expertise for meaningful analysis and interpretation. Not suitable for use w...	discouraged-003	Analysis Without Domain Expertise

License And Use Terms

ID

license-001

Name

Creative Commons Attribution Non-Commercial Share-Alike

Description

Data licensed for reuse under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International license (https://creativecommons.org/licenses/by-nc-sa/4.0/). Attribution is required to the copyright holders and the Cell Maps for Artificial Intelligence project. Any publications referencing this data or derived products should cite the bioRxiv article (Clark T, et al. Cell Maps for Artificial Intelligence: AI-Ready Maps of Human Cell Architecture from Disease-Relevant Cell Lines. BioRXiv, May 2024. doi:10.1101/2024.05.21.589311) and directly cite the data collection. Commercial use requires separate license negotiation with copyright holder (UCSD, Stanford, and/or UCSF depending upon specific data package). A Data Access Committee will supervise ethical matters related to dataset distribution and potential dual licensing for commercial use. Copyright (c) 2025 The Regents of the University of California except where otherwise noted. Spatial proteomics raw image data is copyright (c) 2025 The Board of Trustees of the Leland Stanford Junior University.

License Terms

Attribution required to copyright holders and authors
Non-commercial use only (commercial requires separate license)
Share-alike - derivative works must use same license
Must cite bioRxiv publication and data collection DOI
Data Access Committee oversight for ethical distribution

Motivation

Composition

Collection Process

Uses

Distribution

Maintenance

Human Subjects