Skip to main content
TargetScan Seed Site

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.

Source
evo-design/proto-language/proto_language/constraint/sequence_annotation/targetscan_site_constraint.py
View source
Constraint contributors
Score miRNA seed-match site burden using TargetScan’s canonical site definitions. For each target sequence and each miRNA, canonical seed-match sites are detected and summed with per-site-type and per-miRNA weights into an aggregate burden, bounded by repression_threshold. direction='maximize' rewards installing sites (low score when burden is high); direction='minimize' rewards escaping them.

​
API Reference

ConfigTargetScanSiteConfig Source
Configuration for TargetScan canonical seed-match site scoring.
​
List[string]
required
miRNA query sequences (>= 8 nt) whose seed (positions 2-8) defines target sites.
​
array
Optional identifiers for miRNA queries.
​
array
Optional nonnegative per-miRNA expression weights matching the miRNA query order.
​
enum
default:"maximize"
Whether predicted miRNA site burden should be high or low.Options: maximize, minimize
​
number
default:"2.0"
Aggregated site-burden score treated as saturating success.
​
object
Per-site-type strengths for 8mer/7mer-m8/7mer-A1/6mer.
​
boolean
default:"True"
Whether to count weak 6mer sites (set False to require >= 7mer).
ReturnsConstraintOutput
One result per sequence. score is in [0, 1] (0.0 = objective fully satisfied). metadata carries the aggregated targetscan_repression_score, targetscan_num_sites, and a per-site list with miRNA id, site type, weighted strength, and 1-indexed coordinates.

​
Usage

python

​
Metadata