scanRBP: RNA-protein binding toolkit

scanRBP scores how strongly RNA-binding proteins (RBPs) are predicted to bind along a nucleotide sequence, using a database of pre-trained position weight matrices (PWMs), and draws the result as a heatmap.

pip install scanRBP
scanRBP AAAGCGGCGACTTATTATATCCCCATATATTATATCTTCTTCTCTTATATATAAACCAGAGATAGATGTGTGTGGTGG example1 -heatmap example1

Try it online, no installation

scanRBP is available directly in your browser via expressRNA.org — go to Analyses → New Analysis → scanRBP Analysis.

What it computes

For each RBP in the database, scanRBP calculates a per-position log-odds score: how much more (or less) likely each nucleotide is to come from that RBP's binding motif than from background sequence. Positive values mean "more consistent with the motif than background"; 0 means no preference; negative values mean "less consistent than background." See the biopython PSSM docs for the underlying math.

What's included

• PWM-based scanning — score a sequence (or every sequence in a FASTA file) against ~100+ RBP motifs from the mCross and CISBP-RNA databases in one command. See Scoring sequences.
• Motif search — look up which RBP entries are in the database by name, alias or description. See Motif database & search.
• CLIP-based scoring — score against real binding evidence (BED/bedGraph peaks from eCLIP/iCLIP experiments) instead of a PWM. See CLIP-based scoring.
• Heatmap visualization — clustered PNG/PDF heatmaps of binding scores across proteins and positions, with cumulative plots across many sequences for a single protein.
• Python API — call scanRBP's scoring directly from Python for embedding in other tools (this is how the expressRNA scanRBP analysis works). See Python API.

Where to start

New to scanRBP? Read Installation and then Quick Start to get your first heatmap in under a minute. Everything else in these docs is reference material to dig into as you need it.