rjsicko / nbs_ngs Goto Github PK

For samples with no variants, smaller font or grey or something to visually cue ignore them.

Should just be a matter of adding formatting after this else:

https://github.com/rjsicko/scripts/blob/3092428c989466c2b206cf002ace33fdbc787696/CFTR/NYS_CF_2ND_TIER_VARIANT_REPORT_v1.0.0-alpha.1.html.tmpl#L115

see if we can error check the samplesheet.csv for spaces in sample names.
probably would be best to do the check in function run_ready_demultiplex ()

another 2nd tier job report - "3rd tier worklist"
just a list of sample IDs that have 1 or more 2nd tier variants.

either make a spot on the reports that is 'yes' or 'no' that checks this or don't generate the report and instead display a warning

rest_api/jobs/4834/samples/4845/results/variant-filtered-state
"filters": [
{
"column": "AF",
"filter_value": "0.20",
"field_name": "allele_fraction",
"filter_operation": ">=",
"discard_blanks": false
}

for all job hooks - output a log file specifying metadata (versions, files, time etc.)

For improved reports, I was reluctant to use custom django filters as it's more to maintain on VM upgrades. Alternative possibility = custom post job hook (python) that grabs the info it needs and generates custom reports.

it should be possible to queue up 3rd tier analysis for samples using post job hook on 2nd tier run.

the only tricky part will be ultra-high IRT samples. Possible solutions:

• in Archer Analysis set sample attribute marking the high IRTs. job hook would search for these.

• somewhere upload a list of samples that are high IRT and have the job hook scan this list.

https://gau.ccr.cancer.gov/nextseq_details/

for archer_process_miseq_run.sh, could have it look for SampleSheet.corrected.csv this way that file could be placed in the run folder while the run is still going on, but it would be used for demultiplexing instead of the SampleSheet specified when loading the MiSeq

alternative to <th width="X%">, set widths manually since we break_words at certain number of characters... safer

see if we can make homozygous variants count as 2 on the 'variant count' column.

Originally posted by @rjsicko in #1 (comment)

remove the "Deletion?" field from the SV portion of the report.

Deletion?: {{isoform.is_dna_deletion|yesno}}

I thought this was is it a DNA deletion or duplication, but it is actually Archer's variable for DNA vs RNA SV

QC report, make sample ID red if sample fails qc.

NYS_CF_2ND_TIER_VARIANT_REPORT_v1.0.0-rc.1.html.tmpl
isnt working as expected. it is pulling the information from the VCF as expected, but it isn't rendering the pdf. generates an html report. html might be fine, even prefered, but will require formatting for proper display and printing. either work out the html formatting or get the pdf to render.

chop at least "_L001_R1_001" from Sample ID in reports. preferably chop "S##" too but could get complicated by fact that some are single digit (S1-9) and some are double (S10-S96).

figure out why printing from adobe causes the tables to get cutoff, but printing from the browser looks good.

if we check polyTG job hook on a targeted mutation run, the script should check for this and simply output "polyTG not run, targeted mutation run"

"is_only_report_targeted_mutations": true,

look into if the reports need to be adjusted to only account for dropping AF threshold from 0.300 to 0.200