Skip to content

Reads and molecules

Plots that keep the evidence one molecule at a time, before reads are collapsed into depth, a call or an average per site.

How to choose

Your question Plot Build it with
Does a call hold up in the reads behind it? Read pileup PileupTrack, --pileup
Did one molecule align to several places, and in what order? Split reads SplitReadTrack, --split-reads
Are the methylated sites on the same molecules, or scattered across them? Methylation by molecule BisulfiteTrack, --bisulfite
How many reads crossed each intron? Splice junctions JunctionTrack, --junctions
What did the current look like before basecalling? Nanopore signal SquiggleTrack, --squiggle

Plots

  • Reads packed into rows under a depth profile, a candidate SNV and the reference, coloured by strand, with mismatches painted against the reference, a deletion, an insertion, faded low-quality reads and a count of the reads not shown

    Read pileup Reads placed by their real CIGAR and packed into rows, with mismatches found against the reference and the reads past the row limit counted on the band.

  • Eight molecules each aligned in three segments, both ends at a new insertion site and the middle on the reference IS6110 copy on the reverse strand, with the connectors drawn under the rows, above a depth profile that doubles over that copy

    Split reads One row per molecule and one bar per alignment, joined in the order the molecule ran, with any step back down the reference drawn under the row.

  • Sixteen molecules across the H19/IGF2 imprinting control region, one row each, filled circles for methylated CpGs and open circles for unmethylated ones: some reads are methylated almost throughout and others almost nowhere

    Methylation by molecule One row per read, a filled circle where a site is methylated, an open one where it is not and nothing where the read did not reach, so two alleles show as stripes.

  • Splice junction arcs labelled with read counts above an RNA depth profile, with one junction held back and counted, and per-base model attribution beneath

    Splice junctions One arc per intron, weighted and labelled by the reads that crossed it; arcs sit in lanes, so their height means nothing.

  • Raw nanopore current for one read against sample number, stepping between levels, with the base the basecaller assigned to each stretch written above it

    Nanopore signal Raw current against sample number, an envelope of the extremes when zoomed out and the samples themselves when zoomed in; a move table adds the called bases.

  • Signal and sequence

    After the reads have become depth or a methylated fraction per site.

  • Variation and association

    After the evidence has become a call.

  • File formats

    BAM and SAM and its SA tag, STAR junction tables and Bismark extractor files, and how a CRAM comes in through samtools.

  • Recipes

    Complete programs that stack several tracks into one figure.