From 7c03e48bd4722e39660518b4b67e5ee36dc2f05b Mon Sep 17 00:00:00 2001 From: Robert Sidney Cox Date: Wed, 4 Dec 2024 11:22:43 -0500 Subject: [PATCH] Update website/docs/Pipelines/Optimus_Pipeline/optimus.methods.md Co-authored-by: Elizabeth Kiernan <55763654+ekiernan@users.noreply.github.com> --- website/docs/Pipelines/Optimus_Pipeline/optimus.methods.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/website/docs/Pipelines/Optimus_Pipeline/optimus.methods.md b/website/docs/Pipelines/Optimus_Pipeline/optimus.methods.md index 7543be9961..f15e0f7781 100644 --- a/website/docs/Pipelines/Optimus_Pipeline/optimus.methods.md +++ b/website/docs/Pipelines/Optimus_Pipeline/optimus.methods.md @@ -20,7 +20,7 @@ An example of the pipeline and outputs is available on the [Terra HCA Optimus Pi ## Single-nucleus (sn_rna mode) -Data preprocessing and count matrix construction were performed using the Optimus v7.8.1 pipeline (RRID:SCR_018908). Briefly, FASTQ files were partitioned by barcodes using sctools v0.3.13. The files were then trimmed, aligned, UMI-corrected against the 10x Genomics barcodes whitelist, and converted to a raw count matrix using STAR v2.7.9a. CB correction was performed using the `--soloCBmatchWLtype 1MM_multi` parameter which allowed for multiple matches in the whitelist with 1 mismatched base, used posterior probability calculation to choose one of the matches, added pseudocounts of 1 to all whitelist barcodes, and allowed multi-matching of CBs with N-bases to the whitelist. +Data preprocessing and count matrix construction were performed using the Optimus v7.8.1 pipeline (RRID:SCR_018908). Briefly, FASTQ files were partitioned by barcodes using fastqprocess. The files were then trimmed, aligned, UMI-corrected against the 10x Genomics barcodes whitelist, and converted to a raw count matrix using STARsolo v2.7.11a. CB correction was performed using the `--soloCBmatchWLtype 1MM_multi` parameter. Reads were trimmed using the solo parameter `--clipAdapterType CellRanger4` and `--outFilterScoreMin 30` which matches read trimming performed by CellRanger4. Reads were then aligned to GENCODE mouse (M21) or human (V27) references in unstranded mode. Genes were annotated using the STAR "GeneFull" COUNTING_MODE and UMIs were corrected with the `--soloUMIdedup 1MM_CR`, which uses a directional correction method. The resulting BAM was then used for cell and gene metric correction using the sctools v0.3.13 TagSortBam tool. The STAR TSV outputs for gene counts, features, and barcodes were converted to numpy arrays for downstream h5ad conversion. All cell and gene metrics (alignment, mitochondrial, and other QC metrics) and count matrices were aggregated into a final h5ad-formatted cell-by-gene matrix. The final outputs included the unfiltered h5ad and unfiltered (but tagged) BAM file.