COMPASS is a machine-learning classifier that scores (exogenous miRNA, human gene) pairs for likely targeting. Trained on experimentally-validated human miRNA-target interactions (miRTarBase) with TargetScan-style seed-site features, ViennaRNA duplex MFE, and phyloP conservation. The model is agnostic to the source organism — the bundled reference pools (Arabidopsis, rice, maize, soybean) are provided as convenient defaults for cross-kingdom work, and custom FASTA uploads are accepted.

• Forward — given an exogenous miRNA (or FASTA of several), rank human genes by predicted score.
• Reverse — given one or more human gene symbols, rank exogenous miRNAs from a chosen source-species pool. Use the gene dropdown on the results page to navigate between genes.

Retention policy: submitted COMPASS analyses are automatically deleted 30 days after they are generated. Please download any results you wish to keep before then.

This list is stored only in your browser (no server account, no cookies) — other users and other devices see their own empty list. Jobs themselves stay on the server for 30 days; bookmark the URL on the right-hand side if you want to come back from another device.

Submit an analysis