Scoring sequences

This page covers the full set of scoring and plotting options, beyond the basics in Quick Start.

Limiting to specific proteins

By default scanRBP scores a sequence against every RBP in the database. To restrict to one (or a group matching a search term), use -protein:

scanRBP AAAGCGGCGACTT... example1 -protein TARDBP -heatmap example1

-protein is matched the same way as the search command — a case-insensitive substring match against scan_id, protein name and description — so -protein TARDBP can resolve to more than one scan_id if the protein was profiled in multiple cell lines. Run scanRBP search <term> first if you want to see exactly which entries a given -protein value will include.

Cumulative plots across many sequences

-cumulative changes the output shape: instead of one heatmap per sequence (proteins × positions), it produces a single heatmap for one protein across all sequences in a FASTA file (sequences × positions). This is useful for looking at how one RBP's predicted binding varies across many loci — e.g. all 3'-UTRs of a gene family, or all sequences flanking a set of splice sites.

scanRBP loci.fasta -cumulative -protein TARDBP -heatmap tardbp_binding

• Requires -protein to name the protein(s) to plot (there's no "all proteins, cumulative" mode).
• All sequences in the FASTA file must be the same length — scanRBP exits with an error otherwise.
• Only usable with a FASTA file input, not a single command-line sequence.

Heatmap styling

Heatmaps are only generated when -heatmap <title> is given — without it, scanRBP still writes the .tab.gz score matrix but skips the plot.

Option	Effect
-heatmap title	Generate the heatmap (PNG + PDF), with title used as the plot title.
-annotate	Print the numeric score inside each heatmap cell.
-xlabels	Show the sequence itself along the x-axis (off by default, since long sequences make this unreadable).
-figsize "(10,20)"	Set the matplotlib figure size (width, height in inches).
-fontscale 0.2	Scale all heatmap text (default 0.2 — the database has many proteins, so labels start small).
-vlines "10,-5"	Draw dashed vertical guide lines at the given x positions. Negative values count back from the end of the sequence (e.g. -5 is 5 bases before the end) — handy for marking a splice site or crosslink position.

Other options

Option	Effect
-nonzero	Clip all negative scores to 0 before scoring/plotting (off by default — negative scores, meaning "less consistent with the motif than background," are kept).
-output_folder folder	Write all output files here instead of the current directory (created if it doesn't exist).
-force	Overwrite existing output files. Without it, scanRBP skips a sequence whose .tab.gz output already exists — useful for resuming a large FASTA batch.

Output files

For an output name example1 (or a FASTA sequence id seq1, prefixed pwm_):

example1.tab.gz   # log-odds score matrix (proteins x positions), gzipped TSV
example1.png      # heatmap, if -heatmap was given
example1.pdf      # heatmap, if -heatmap was given

+ and - characters in the output name are replaced with plus/minus (common when a FASTA id encodes a genomic strand), so filenames stay filesystem-safe.