Skip to main content
AlphaGenome Interval Track
License: AlphaGenome uses Apache-2.0 for code and Custom (AlphaGenome Terms of Use) for model weights and has restrictions around commercial use and may require explicit attribution when utilized. Model weights are gated and require accepting the provider’s terms and authenticating with a HuggingFace token. Please refer to the code license and model weights license for full terms.

This constraint is open source. Any third-party models, product names, or trademarks referenced are the property of their respective owners, and Proto is not affiliated with them.

Go to Tool Page
evo-design/proto-language/proto_language/constraint/sequence_annotation/alphagenome_interval_track_constraint.py
View source
Copy citation
Constraint contributors
Score mean AlphaGenome track signal over one or more intervals.

​
API Reference

ConfigAlphaGenomeIntervalTrackConfig Source
Configuration for interval track-value scoring with AlphaGenome.
​
List[array]
required
0-based half-open intervals scored relative to the target segment.
​
string
default:""
DNA context placed upstream of the target segment before AlphaGenome prediction.
​
string
default:""
DNA context placed downstream of the target segment before AlphaGenome prediction.
​
integer
default:"16384"
Full sequence length submitted to AlphaGenome after context composition.
​
List[string]
required
AlphaGenome ontology term(s) to score.
​
enum
default:"RNA_SEQ"
AlphaGenome output type to score.Options: ATAC, CAGE, DNASE, RNA_SEQ, CHIP_HISTONE, CHIP_TF, SPLICE_SITES, SPLICE_SITE_USAGE, SPLICE_JUNCTIONS, CONTACT_MAPS, PROCAP
​
array
Keywords selecting individual tracks (e.g. [‘H3K4me1’]) from a bundled output; None = all tracks.
​
enum
default:"maximize"
Whether to maximize or minimize the interval mean signal.Options: maximize, minimize
​
number
default:"5.0"
Signal value where maximize-mode sigmoid has its inflection.
​
number
default:"1.0"
Scale for maximize-mode sigmoid transform.
​
number
default:"1.0"
Upper bound for minimize-mode linear score; values above are clipped.
​
array
Optional off-target cell-type term(s); when set, scores the target-vs-off-target signal margin.
​
number
default:"0.0"
Signal-difference (target - off-target) where the contrastive sigmoid has its inflection.
​
number
default:"1.0"
Scale for the contrastive-mode margin sigmoid transform.
​
string
default:"all_folds"
AlphaGenome model version.
​
enum
default:"human"
Organism for AlphaGenome prediction.Options: human, mouse
​
string
default:"cuda"
Device for AlphaGenome prediction.
​
integer
default:"3600"
Timeout (seconds) for each AlphaGenome prediction call.

​
Usage

python

​
Metadata