point of discussion for pairtools module

open2c / MultiQC

Aggregate results from bioinformatics analyses across many samples into a single report.

http://multiqc.info

GNU General Public License v3.0

2 stars 2 forks source link

point of discussion for pairtools module #2

Open sergpolly opened 2 years ago

sergpolly commented 2 years ago

chromsizes in the stats reports - yes/no ?!
how "rigid" are our cis-ranges (<1kb, 1-2kb, ...>40kb) - should we skip analysis when different ones encountered ?
interactions by chromosome pairs (the last heatmap) - remove it for now ? how to display in the future ?

sergpolly commented 2 years ago

more points dug out of our slack:

%cis in the main table is confusing ( @mimakaev @gfudenberg proposed other way to calculate it - couldn't find now )
@gfudenberg proposed to add self-circles and dangling ends to main table - are we going to over-interpret FR RF type of pairs for that ? e.g. what's the definition of a dangling end - range <1 kb and FR orientation ? why not < 2kb ? microC vs Hi-C ? should we just name them something else e.g. "short-range FR(dangling end enriched)" ?

sergpolly commented 2 years ago

how else can we recycle pairwise chromosomal counts - ideas ? right now we try to plot these counts in a form of a heatmap with rows as samples and columns as counts of inter-chromosomal interactions (sorted and truncated at some arbitrary point)

can we condense it into something smaller but still useful - e.g. cis/trans on a per chromosome level ? maybe per-chromosome coverage (this would require chromsizes though) ?

yeah - I guess ideally one would infer copy-number variations and potential transolcations from this ...