-
Notifications
You must be signed in to change notification settings - Fork 2
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add process for building cellranger index to build-index.nf #66
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This looks good. I made a couple of suggestions below, which might make this index more comparable to the one we use for salmon.
I was going to make an additional suggestion about naming params, but ended up putting it in a separate issue: #67
build-index.nf
Outdated
cellranger mkgtf \ | ||
genome.gtf \ | ||
filtered.gtf \ | ||
--attribute=gene_biotype:protein_coding |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Is this going to be too stringent? I would think we should include ncRNA and such: The default cellranger index includes much more: https://support.10xgenomics.com/single-cell-gene-expression/software/pipelines/latest/advanced/references#mkgtf
(I just looked at the full list of gene_biotypes
in the reference we are using, which appears below (generated with grep -o 'gene_biotype [^;]*' Homo_sapiens.GRCh38.104.gtf |sort | uniq
) and noted that the code used by 10X might miss lncRNA
as it is looking for lincRNA
, according to that page.)
Alternatively, we could probably just not filter the gtf file? This would probably be more comparable to what we have for the salmon index, which I do not think we filter at all (hence 60K "genes").
gene_biotype "IG_C_gene"
gene_biotype "IG_C_pseudogene"
gene_biotype "IG_D_gene"
gene_biotype "IG_J_gene"
gene_biotype "IG_J_pseudogene"
gene_biotype "IG_V_gene"
gene_biotype "IG_V_pseudogene"
gene_biotype "IG_pseudogene"
gene_biotype "Mt_rRNA"
gene_biotype "Mt_tRNA"
gene_biotype "TEC"
gene_biotype "TR_C_gene"
gene_biotype "TR_D_gene"
gene_biotype "TR_J_gene"
gene_biotype "TR_J_pseudogene"
gene_biotype "TR_V_gene"
gene_biotype "TR_V_pseudogene"
gene_biotype "lncRNA"
gene_biotype "miRNA"
gene_biotype "misc_RNA"
gene_biotype "polymorphic_pseudogene"
gene_biotype "processed_pseudogene"
gene_biotype "protein_coding"
gene_biotype "pseudogene"
gene_biotype "rRNA"
gene_biotype "rRNA_pseudogene"
gene_biotype "ribozyme"
gene_biotype "sRNA"
gene_biotype "scRNA"
gene_biotype "scaRNA"
gene_biotype "snRNA"
gene_biotype "snoRNA"
gene_biotype "transcribed_processed_pseudogene"
gene_biotype "transcribed_unitary_pseudogene"
gene_biotype "transcribed_unprocessed_pseudogene"
gene_biotype "translated_processed_pseudogene"
gene_biotype "translated_unprocessed_pseudogene"
gene_biotype "unitary_pseudogene"
gene_biotype "unprocessed_pseudogene"
gene_biotype "vault_RNA"
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We definitely don't do any filtering like this for the salmon index, so I can remove this to make it more comparable.
Co-authored-by: Joshua Shapiro <josh.shapiro@ccdatalab.org>
I went ahead and removed filtering and it worked fine, creating a larger index this time. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good, with two small suggestions I made, and one that I can't actually put in, which is to add
cellranger_index = "s3://nextflow-ccdl-data/reference/homo_sapiens/ensembl-104/cellranger_index/Homo_sapiens.GRCh38.104_cellranger_full"
to the params in nextflow.config
build-index.nf
Outdated
cellranger mkgtf \ | ||
genome.gtf \ | ||
filtered.gtf | ||
|
||
cellranger mkref \ | ||
--genome=${cellranger_index} \ | ||
--fasta=genome.fasta \ | ||
--genes=filtered.gtf \ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think you can actually completely skip mkgtf
as I don't think it is doing anything here.
cellranger mkgtf \ | |
genome.gtf \ | |
filtered.gtf | |
cellranger mkref \ | |
--genome=${cellranger_index} \ | |
--fasta=genome.fasta \ | |
--genes=filtered.gtf \ | |
cellranger mkref \ | |
--genome=${cellranger_index} \ | |
--fasta=genome.fasta \ | |
--genes=genome.gtf \ |
Co-authored-by: Joshua Shapiro <josh.shapiro@ccdatalab.org>
@jashapiro I went ahead and made the edits that you suggested and then also had to increase the memory that the process is using. I noticed that it was only making some of the index files but not all of them even though it was completing the process successfully without error. It wasn't until I attempted to run spaceranger using the index later on that I realized there was missing files and only a partial index so had to bump up the memory required too. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM!
Closes #65. This PR adds the
cellranger_index
process to the workflow for building the indices,build-index.nf
ensuring that when we build all of our indices they will use the same assembly. I followed the same setup that was previously used in generating the cellranger index inalsf-scpca/workflows/rnaseq-ref-index/build-cellranger-index.nf
.This was pretty straight forward and I was able to test it and everything ran successfully, producing a new index with ensembl 104. The only question I had was about how we wanted to name this index. I included the assembly name that we use in the splici index and then appended
_cdna_cellranger
, but let me know if there is a different naming convention that would be preferred.