{"endpoint": "POST /api/v1/batch/analyze/", "description": "Batch cohort analysis over many proteins/regions; same engine as the Batch Analysis page.", "content_type": "application/json", "documentation": "/api_documentation#batch-analysis-api", "note": "Send JSON with a \"roi\" string (newline-separated identifiers, optional start end per line).", "search_type": "gene_name | uniprot | gencode (Gencode protein accession, same as Summary URL)", "go_gene_list": "GET /ajax/go-genes/?id=GO:0005813 returns main-isoform accessions + symbols (aligned with Statistics dynamic GO).", "ppi_neighbourhood": "GET /ajax/ppi-proteins/?hub=SIAH1 returns hub + direct partners as main-isoform accessions (Statistics PPI logic).", "response_extras": "plot_payload (Plotly-friendly series), motif_visualization (PSSM matrix or regex sequences for logos).", "pssm_tsv": "Use pssm_input_format: \"tsv\" and pssm_tsv multiline string (optional header row of one-letter AAs).", "island_filter": "island_pathogenicity (bool), island_metric: alphamissense | conservation, island_conservation_field: phastcons | global, island_conservation_delta, island_conservation_core_min, island_flank_mode: motif | fixed, island_flank (fixed mode), island_flank_aggregate: max | mean, island_score_delta, island_am_mean.", "dbnsfp_columns": "Optional output keys \u2014 region mean of a dbNSFP 5.x VEP **rankscore**. One key per predictor as patho_<name>, e.g. patho_sift, patho_revel, patho_clinpred, patho_esm1b, patho_alphamissense, patho_cadd, patho_popeve. All 40 names are listed in DBNSFP_PREDICTOR_LABELS. Rankscores run 0-1 with higher = more damaging for every tool, including SIFT and PROVEAN whose raw scales are inverted \u2014 so means are comparable across tools, but not with values produced before the dbNSFP 5.x migration. patho_eve is no longer available (dbNSFP 5.3.1a has no EVE column); use patho_popeve instead.", "find_region": "Annotation: Elm, PEM, Mobidb, PTM, Significantly_Mutated, Pfam, \u2026 \u00b7 Prediction: regex / PSSM", "ptm_type_filter": "With find_region=PTM: filter by Type substring (e.g. Phosphorylation, Acetylation)"}