Figure: CRISPRs

• make plots/table to represent the CRISPR spacer identification

P.S. while the files are ready, I still have to create snakefiles. Will do later today and let you know!

changed milestone to %Manuscript - initial version

added Figure label

mentioned in issue #24 (closed)

• output files after this commit-e98663ed, will be under results/analysis/crispr/{folder}
• folder == ["casc", "minced"]

Notes

Files to parse

• casc: <tool>.results.txt
• minced: <tool>.txt

Information

• contig ID
• array start/end

Plot

• intersection plot: UpSetR, contig IDs only
• overlap plot: array overlap per contig

mentioned in commit 629dc3fc

mentioned in commit d75c0381

First attempt: fig_crispr.pdf

I am not happy with these plots...

Problems:

• An intersection plot over the contig IDs does not tell much
  • Also, saving multiple UpSetR plots is problematic for some reason
• Could not find a good way to plot CRISPR array coordinates
  • Too many contigs to plot them individually
  • Plotting contigs together requires a large coordinate range s.t. the arrays cannot be seen properly

Any suggestions are welcome. :)

• maybe do a simple facet? the overlap is a great idea but not critical.

What do you mean by "a simple facet"? To plot what?

• facet the tools.. keep them separate..
• within each tool, compare the different assembly methods ?!

I mean what exactly should I compare?

Currently, I only look at the overlap between minced and crisp based on contig IDs. I cannot compare between assemblies as the contig IDs are different.

My bad.. I should have explained better.

• we simply report the number of spacers recovered, i.e. something like this: less minced_CRISPR_output.txt

flye
44
megahit
40
metaspades_hybrid
169
metaspades
192

• but we normalise that to total number of contigs/total length of sequences?!
• could be a barplot, or a mere table even.
• And we do this separately for CASC and minCED
• Since the overlap is a pain anyways to represent graphically

changed title from figure-request-5: CRISPRs to Figure: CRISPRs

mentioned in commit 1990f776

mentioned in commit 25b450c3

Plot update: fig_crispr.pdf

Notes about how the data is collected:

• CasC: reading in {asm_tool}.results.txt
• MinCED: parsing {asm_tool}.txt
  • for each sequence and each found CRISPR array: extract the spacer sequences
    • number of spacers: number of unique spacer sequences

Created summary contains for each CRISPR tool and assembly the found CRISPR arrays w/ sequence (contig) ID, array start/end and the number of spacers.

closed

Filter CasC table by bonafide.

reopened

mentioned in commit a12b69cc

fig_crispr.pdf

closed