Chrombpnet Tools¶
Configuration File: remote_tools/chrombpnet_tools.json
Tool Type: Remote
Tools Count: 2
This page contains all tools defined in the chrombpnet_tools.json configuration file.
Available Tools¶
run_chrombpnet_predict (Type: RemoteTool)¶
Predict chromatin accessibility (ATAC/DNase) from a DNA sequence with a trained ChromBPNet model …
run_chrombpnet_predict tool specification
Tool Information:
Name:
run_chrombpnet_predictType:
RemoteToolDescription: Predict chromatin accessibility (ATAC/DNase) from a DNA sequence with a trained ChromBPNet model (Pampari 2025) — the bias-corrected, base-resolution successor to DeepSEA/Basset. The sequence is center-cropped/N-padded to 2,114 bp; returns predicted log total counts (coverage magnitude), total counts, and the base-resolution accessibility profile (1,000 bp). Requires a server-side cell-type-specific model_path (.h5).
Parameters:
model_path(string) (required) Server-accessible path to a trained ChromBPNet Keras model (.h5), cell-type-specific.sequence(string) (required) DNA sequence (A/C/G/T/N); cropped/padded to 2,114 bp around its center.return_profile(boolean) (optional) Include the full 1,000-bp profile array in the response (default false; summary stats always returned).
Example Usage:
query = {
"name": "run_chrombpnet_predict",
"arguments": {
"model_path": "example_value",
"sequence": "example_value"
}
}
result = tu.run(query)
run_chrombpnet_variant_effect (Type: RemoteTool)¶
Score a non-coding regulatory variant with ChromBPNet (Pampari 2025): predict accessibility for t…
run_chrombpnet_variant_effect tool specification
Tool Information:
Name:
run_chrombpnet_variant_effectType:
RemoteToolDescription: Score a non-coding regulatory variant with ChromBPNet (Pampari 2025): predict accessibility for the reference and alternate sequences (each 2,114 bp, variant at center) and return the count log2 fold-change (alt vs ref accessibility magnitude) and the profile Jensen-Shannon divergence (change in base-resolution accessibility shape) — the canonical ChromBPNet variant-effect scores for interpreting non-coding GWAS/eQTL variants. Requires a server-side cell-type-specific model_path (.h5).
Parameters:
model_path(string) (required) Server-accessible path to a trained ChromBPNet Keras model (.h5), cell-type-specific.ref_sequence(string) (required) Reference DNA sequence centered on the variant (cropped/padded to 2,114 bp).alt_sequence(string) (required) Alternate DNA sequence (same length/centering as ref).
Example Usage:
query = {
"name": "run_chrombpnet_variant_effect",
"arguments": {
"model_path": "example_value",
"ref_sequence": "example_value",
"alt_sequence": "example_value"
}
}
result = tu.run(query)