Append Bjorn output csv and indels to full-summary file

[kmfisch]

From: "Laurent, Louise"
Date: Thursday, March 18, 2021 at 11:21 AM
Subject: Re: Indel flag and Pangolin calls

Hi All,
A few adjustments to the below, based on Kristian's input on this morning's 10 am call -- pending Katie/Amanda checking with Karthik and Al that these fit what Bjorn does:

[snip]

2. Bjorn outputs A new manual_review_flagged column [strike-out original] will be added to the summary csv files. Any sample whose id is returned by line 40 of the github gkarthik/tabulate_variants.py script will have its manual_review_flagged value set to TRUE. All other samples will have their manual_review_flagged value set to FALSE. ​UCSD folks to "observe" Al/Karthik running Bjorn to understand nitty gritty of steps/settings.

[snip]

[AmandaBirmingham]

CCBB has been provided 3 "bjorn outputs": a deletions.csv file, an insertions.csv file, and a "flagged_by_bjorn.txt" file. The last of these appears to be what is referred to above as the return value of line 40, and can reasonably be merged into the *summary.csv files. The deletions.csv example we have been provided appears to be per-contiguous-deletion in a sample's consensus sequence, not per sample. Assuming this is true and that the same is true for insertions.csv, merging these data into the per-sample *summary.csv file would be messy, difficult, and potentially make the file considerably less usable.

[AmandaBirmingham]

Stakeholder: should we do the requested merge of deletions.csv and insertions.csv?

[rob-knight]

Current plan is that Andersen lab will run bjorn and be responsible for merge. This may be revisited but it is low priority and does not block release of the first datasets.