RNA-seq issueshttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues2021-07-05T17:54:39-05:00https://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/130Remove hacky TIN report in MultiQC2021-07-05T17:54:39-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduRemove hacky TIN report in MultiQCMultiQC v1.11 is out with compatibility to report RSeQC TIN values, so ghenry's hacky method of reporting it isn't necessary anymore...MultiQC v1.11 is out with compatibility to report RSeQC TIN values, so ghenry's hacky method of reporting it isn't necessary anymore...v2.x.xhttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/128Check if alignData to depend on checkMetadata errors == false2021-07-15T16:01:10-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduCheck if alignData to depend on checkMetadata errors == falseMemory serves it alignData should require checkMetadata errors == false... but it doesn't appear to be that way now... double check!Memory serves it alignData should require checkMetadata errors == false... but it doesn't appear to be that way now... double check!v2.x.xhttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/126Move uploadInputBag to occur right before uploadExecutionRun2021-07-15T16:01:13-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduMove uploadInputBag to occur right before uploadExecutionRunCurrently, upload of input bag back to hub only depends on fetchBag process... so it is easy to create orphan (without associated execution run entries) inputBag entries in the event of pipeline/cluster failures.
Making uploading of inp...Currently, upload of input bag back to hub only depends on fetchBag process... so it is easy to create orphan (without associated execution run entries) inputBag entries in the event of pipeline/cluster failures.
Making uploading of input bag dependent speciesInfer will prevent it from happing until just before the upload execution run process.
(This is from a conversation with hongsuda 202010421v2.x.xhttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/124Some seqwho seq-type inferences have commas which mess up error descriptions2021-07-15T16:01:16-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduSome seqwho seq-type inferences have commas which mess up error descriptions# Summary
chip,dnase seq-type inference from seqwho's comma messes with execution run status description output to trunkate
# Steps to reproduce
run Q-Y5E6
# Observed bug behavior
**Infered sequencing type does not match for R1 and R2:...# Summary
chip,dnase seq-type inference from seqwho's comma messes with execution run status description output to trunkate
# Steps to reproduce
run Q-Y5E6
# Observed bug behavior
**Infered sequencing type does not match for R1 and R2:** Infered R1 = chip and infered R2 = chip
# Expected behavior
**Infered sequencing type does not match for R1 and R2:** Infered R1 = chip and infered R2 = chip,dnase
# Relevant logs and/or screenshots
# Potential fixes
20210420:\
cbennett 9:50 PM\
@Gervaise no that's all that will output with a comma. I can change that for you to be something else?\
Gervaise:face_with_cowboy_hat: 9:51 PM\
...that would be awesome for me... then I wouldn't have to modify my pipeline... just the docker container... hehe\
cbennett 9:51 PM\
Yeah no worries. I'll push it out as a minor update\
Gervaise:face_with_cowboy_hat: 9:51 PM\
awesome... let me know when...\
Venkat Malladi 9:52 PM\
I missed it with such a big code merge\
9:52\
Sorry @Gervaise\
cbennett 9:52 PM\
I've added a reminder to my calendar. Expect end of week\
**ONCE SEQWHO VERSIONED (EXPECT BY 20210425) VERSION DOCKER CONTAINER AND CAN USE IN PIPELINE**
/cc @ghenry @venkat.malladi @s181706v2.x.xhttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/123Handle image caching in CI better2021-03-05T11:07:23-06:00Gervaise Henrygervaise.henry@utsouthwestern.eduHandle image caching in CI better`cat nextflow.config | grep -oP "container = \K.*" | tr -d "'" | sort | uniq | xargs -P 3 -I {} bash -c "singularity pull --dir ${dir} 'docker://'{} || true"`
assumes:
- containers referred in nextflow.config are on docker hub and do not...`cat nextflow.config | grep -oP "container = \K.*" | tr -d "'" | sort | uniq | xargs -P 3 -I {} bash -c "singularity pull --dir ${dir} 'docker://'{} || true"`
assumes:
- containers referred in nextflow.config are on docker hub and do not contain the "docker://" prefix
- containers referred in nextflow.config are prefixed by "container = " **EXACTLY** if a space is missing for example, that container will be skipped
need to make this more robustFUTURE Versionhttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/117Update $31 snippet to use bash opt-args2021-03-19T18:05:50-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduUpdate $31 snippet to use bash opt-argsinstead of numbered inputsinstead of numbered inputsFUTURE VersionVenkat MalladiVenkat Malladihttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/94Update tracking db with on error2021-07-05T12:24:18-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduUpdate tracking db with on errorFUTURE Versionhttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/93Add option for long lived globus-credential file2021-07-05T12:25:17-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduAdd option for long lived globus-credential filein addition to current cookie and credential files
re Laura & Mike, servers set to require separate long-lived tokens, and deriva only keeps 1 long-lived token active
they are considering allowing multiple tokens to be active and if de...in addition to current cookie and credential files
re Laura & Mike, servers set to require separate long-lived tokens, and deriva only keeps 1 long-lived token active
they are considering allowing multiple tokens to be active and if decide to accept will be in another deriva update.
if they don't then long-lived token will work for production only runsFUTURE Versionhttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/59Add new GRCm392021-04-20T21:59:50-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduAdd new GRCm39FUTURE VersionVenkat MalladiVenkat Malladihttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/58AWS/DNANeuxs Cost2021-02-22T17:28:14-06:00Gervaise Henrygervaise.henry@utsouthwestern.eduAWS/DNANeuxs Cost*Currently in GUDMAP/RBK (20200723)*
607 current replicates
497 legacy "samples" (equivalent of replicates)
They want to migrate the legacy samples to be replicates, I don't know how much that has been done
worst case scenario is that 1...*Currently in GUDMAP/RBK (20200723)*
607 current replicates
497 legacy "samples" (equivalent of replicates)
They want to migrate the legacy samples to be replicates, I don't know how much that has been done
worst case scenario is that 1104 replicates to run...
20200722 test using AWS spot instances cost ~$0.984 for 1 PE replicate
20200723 test using AWS spot instances cost ~$0.012 for 1 PE replicate
***NEED TO TEST RUN WITH ON-DEMAND INSTANCES TO SEE WHAT MAX COST PER REPLICATE IS***FUTURE Versionhttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/55Possibly move params.source to config?2021-07-05T12:22:52-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduPossibly move params.source to config?FUTURE Versionhttps://git.biohpc.swmed.edu/gudmap_rbk/rna-seq/-/issues/66On process fail retry2021-04-20T21:58:32-05:00Gervaise Henrygervaise.henry@utsouthwestern.eduOn process fail retrypossibly with bigger nodepossibly with bigger node