Predictive Models

Coding variant effects/prioritizationNon-coding and coding variant effects/prioritizationCADDDeleteriousness of variantsFile of variants, precomputation of complete hg37/38 availablehttps://cadd.kircherlab.bihealth.org/Martin Kircher, BIH
Coding variant effects/prioritizationImpact on functional residues, e.g. protein-protein interaction residues, stability, PTM sites, etc.MutPred2A general score indicative of pathogenicity and a ranked list of impacted mechanisms (scores and P-values)Protein sequences in FASTA format with amino acid variants in XnY formathttp://mutpred.mutdb.org/Predrag Radivojac, Northeastern
Coding variant effects/prioritizationImpact on functional residues, e.g. protein-protein interaction residues, stability, PTM sites, etc.MutPred-LOFA general score indicative of pathogenicity and a ranked list of impacted mechanisms (scores and P-values)Protein sequences in FASTA format with amino acid variants in ANNOVAR formathttp://mutpred2.mutdb.org/mutpredlof/Predrag Radivojac, Northeastern
Coding variant effects/prioritizationImpact on functional residues, e.g. protein-protein interaction residues, stability, PTM sites, etc.MutPred-IndelA general score indicative of pathogenicity and a ranked list of impacted mechanisms (scores and P-values)Protein sequences in FASTA format with amino acid variants in ANNOVAR formathttp://mutpred2.mutdb.org/mutpredindel/Predrag Radivojac, Northeastern
Gene Regulatory NetworksDREMDynamic regultory map of gene expression bifurcations and TFs associated with themTime series gene expression and (static) TF-gene associationshttps://github.com/jernst98/STEM_DREMhttp://sb.cs.cmu.edu/drem/Jason Ernst, UCLA
Gene Regulatory NetworkscNMFSets of genes that are cofunctional/coexpressed in a given cell typesingle-cell RNA data from tissue or Perturb-seqhttps://github.com/dylkot/cNMFJesse Engreitz, Stanford
Noncoding variant effects/prioritizationGWASPoPSCausal gene in a GWAS locusSummary stats, catalog of gene setshttps://pmc.ncbi.nlm.nih.gov/articles/PMC10836580/Hilary Finucane, Broad & Jesse Engreitz, Stanford
Noncoding variant effects/prioritizationGWASABC-MaxCausal gene ( with variant and cell context) in a GWAS locus; cell types enriched for fine-mapped variants for a traitscATAC-seq/ATAC-seq/H3K27ac + fine-mapped variantshttps://pubmed.ncbi.nlm.nih.gov/33828297/Jesse Engreitz, Stanford
Noncoding variant effects/prioritizationTF binding, regulatory element annotationBPNetEffect of a sequence variant on signal counts/shape in various assays e.g. ChIP-seqA single epigenomic dataset (e.g. ChIP-seq)https://github.com/kundajelab/bpnetAnshul Kundaje, Stanford
Noncoding variant effects/prioritizationTF bindingQbicCalculates binding affinity change of a sequence variant (from reference) for a given TFfile of variantshttp://qbic.genome.duke.eduAndrew Allen, Duke & Raluca Gordan, Duke
Noncoding variant effects/prioritizationReMMDeleteriousness of variantsFile of variants, precomputation of complete hg37/38 availablehttps://remm.kircherlab.bihealth.org/Martin Kircher, BIH
Noncoding variant effects/prioritizationCTMCdisease-specific (and optionally cell type-specific) variant-gene pairseQTL, GWAS, known variant-gene links, ATAC-seq, variant-TFBSs, GO and expression-based gene-gene similaritieshttps://academic.oup.com/bib/article/22/2/2161/5809565Maureen Sartor, UMich
Noncoding variant effects/prioritizationSURF and TURFVariant effect on TF binding in cell specific mannerDNase-seq, TF & Histone ChIP-seq, Footprints, PWMs through RegulomeDBhttps://regulomedb.org/regulome-search/Alan Boyle, UMich
Noncoding variant effects/prioritizationcoding variants and gene annotationsFAVORFunctionality of variants9 billion variants, multi-faceted pre-collected variant and gene functional annotationshttps://favor.genohub.org/Xihong Lin, HSPH
Noncoding variant effects/prioritizationDragoNNFruitEffect of a sequence variant on signal counts/shape in various single-cell and multimodal assays e.g. scATAC-seqNucleotide sequence and cell representations from a single-cell/multiome experimenthttps://github.com/jmschrei/dragonnfruitAnshul Kundaje, Stanford
Primary categorySecondary categoriesNamePrediction OutputPrediction InputURLLabs
Regulatory element annotationChromHMMChromatin state annotations, cell type specific or cross-cell typeChIP-seq, ATAC/DNasehttps://ernstlab.biolchem.ucla.edu/ChromHMM/https://www.nature.com/articles/nmeth.1906Jason Ernst, UCLA
Regulatory element annotationENCODE cCREsAnnotated promoter, enhancers, and other regulatory elements across hundreds of cell and tissue typesDNase/ATAC, H3K4me3, H3K27ac, CTCFscreen.encodeproject.orgZhiping Weng, UMass & Jill Moore, UMass
Regulatory element annotationCAPRACalculates per element characterization score for WG-STARR-seq data and allow for studying combinatorial effectsWG-STARR-seq + element listhttps://github.com/Moore-Lab-UMass/CAPRAJill Moore, UMass
Regulatory element annotationMACS2Does a variant overlap a DNase or ATAC peak (e.g., extended to a constant distance threshold)ATAC-seq (bulk or sc), DNase-seqhttps://github.com/macs3-project/MACS
TF binding/motif discoveryNoncoding variant effectsZMotifLocations of high confidenceTF motifsTF ChIP-seq, DNase/ATAC-seqhttps://github.com/weng-lab/ZMotifZhiping Weng, UMass
TF binding/motif discoveryTF footprinting from ATAC dataPRINTPredicting TF binding (footprints) from ATAC dataATAC-seq (bulk or sc)https://github.com/buenrostrolab/PRINTJason Buenrostro, Broad
Target gene prediction (element)ABCPercent effect of DNase peak on gene expression in a given cell type; and whether a given ATAC/DNase peak is predicted to regulate any gene in a given cell typescATAC (minimal). Ideally can also use H3k27ac ChIP-seq, Hi-C, good TSS annotationshttps://github.com/broadinstitute/ABC-Enhancer-Gene-PredictionJesse Engreitz, Stanford
Target gene prediction (element)SCENTEnhancer-gene links based on single-cell multimodal dataSingle-cell multiome (ATAC+RNA) datahttps://www.medrxiv.org/content/10.1101/2022.10.27.22281574v1Soumya Raychaudhuri, Brigham and Women’s Hospital
Target gene prediction (element)ENCODE-E2GEnhancer-gene regulatory connections predicted from DNase-seq derived features in 352 ENCODE4 biosamples from a logistic regression model trained on K562 CRISPRi enhancer screen data.DNase-seq, genome annotations/featureshttps://github.com/karbalayghareh/ENCODE-E2GChristina Leslie, MSKCC & Jesse Engreitz, Stanford
Target gene prediction (variant)cS2GTarget gene for every variantGenome-wide SNP-gene predictionshttps://www.nature.com/articles/s41588-022-01087-y and data in https://alkesgroup.broadinstitute.org/cS2G/Steven Gazal, USC
To see more Models, review our Model Set collection