Identify example samples

lcolladotor commented 4 years ago

Add a directory and a Rmd or R file with the code for selecting around 40 RNA-seq samples from one of the public LIBD projects (BrainSeq Phase I, II or maybe even III; stem cell; DG; BPD, ...) with:

Some samples that are part of a complex problem that would need to be dropped unless re-genotyping was an option (not for the example scenario)
Some samples that are part of a simpler problem that can be un-swapped (pending confirmation by re-genotyping; they'll be used in the example scenario).

So after dropping the first set of samples (complex problem) we want a balanced scenario of samples across diagnosis status, age (similar mean/median & range), and sex.

Use the main LIBD RSE with ~5,500 samples to find the JHPCE paths to the FASTQ files for the selected samples.

Main output

Provide a small table (Rdata maybe or a csv file) with:

RNum
BrNum
SwappingCase: no_swap, swap_simple, swap_complex
FASTQpath: could be more than one per sample which you can store as a S4Vectors::CharacterList() in a S4Vectors::DataFrame() object.
Age
Sex
Primary diagnosis
Brain region (ideally all samples should be from 1 brain region, though this will depend on what LIBD project you choose)
RIN

lahuuki commented 4 years ago

issues 5+9 and 10+100+101 are all from psychENCODE_BP may be good candidates

lahuuki commented 4 years ago

Breakdown of all single dataset issues: Nicotine_NAc: 12 psychENCODE_MDD: 13,15,16,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,85,86,87,88 psychENCODE_BP: 5,9,10,100,101 BrainSeq_Phase3_Caudate: 3,4 BrainSeq_Phase1: 52,53,91,92,93,96,97,98,99

lcolladotor commented 4 years ago

The fastq info can be saved as a samples.manifest file like the ones used by SPEAQeasy (@Nick-Eagles can help if needed). That is, subset /dcl01/lieber/ajaffe/lab/zandiHyde_bipolar_rnaseq/preprocessed_data/.samples_unmerged.manifest with the correct paths (from Nina and/or Andrew). So no need to add FASTQpath anymore to the subset of colData(rse) (save this small file and commit it).

lcolladotor commented 4 years ago

For writing the manifest file, use write.table(new_manifest, file = opt$sampleids, row.names = FALSE, col.names = FALSE, quote = FALSE, sep = '\t') https://github.com/LieberInstitute/RNAseq-pipeline/blob/ab71dedb36bcc3dad57233e645fabd5deb96d446/sh/find_sample_info.R#L91

lcolladotor commented 3 years ago

@lahuuki can you verify whether it's ok to close this issue? Thanks!

lahuuki commented 3 years ago

We can close this, we finished up that example a while ago.

LieberInstitute / SPEAQeasy-example

Identify example samples #1

Main output