Skip to content

single cell variant calling

Jump to bottom
Brian Haas edited this page Mar 13, 2024 · 9 revisions

Calling Variants from Single Cell Transcriptomes

Our current approach for detecting variants and mutations in single cells involves running CTAT-Mutations on all reads, identifying variants, and then identifying the cells expressing reference or alternate allelic variants.

Running CTAT-Mutations with single cell transcriptome data

Before running single cell RNA-seq through CTAT-Mutations, the names of the reads should be encoded with cell barcode and UMI information in the following format:

cellbarcode^UMI^read_name

If you have 10xGenomics reads in a ubam format, you can convert to fastq format with the above read name encoding using this script: 10x_ubam_to_fastq.py

Then, run CTAT-Mutations using this fastq file as input along with the command-line flag '--is_single_cells'

Single Cell Outputs

When run with parameter '--is_single_cells', an additional output file will be generated: '${sample_name}.annot_pass_reads.vcf.sc_reads.gz' with the following format:

chr_pos_variant num_reads_with_variant  reads_with_variant      num_ref_matching_reads  ref_matching_reads
chr1:14436:G:A  25      TTCGAAGTCACGACTA^GACCAGGGTTTCCCACCAAC^molecule/84621990,ATTATCCCAATCACAC^TTTCGGCCAATTTCTTATAT^molecule/18857150,...     271     GTAGTCACAGAAGCAC^TTGTCCTGCGTTTCTTATAT^molecule/63537222,ATCATCTTCCAGATCA^TTTACATGCCTTTCTTATAT^molecule/16809660,...
...

which includes, in tab-delimited format:

  • the variant found, formatted as: chromosome:position:REF:ALT
  • number of reads that contain the variant
  • identity of the reads (comma-delimited) that contain the variant
  • number of reads that match the reference allele
  • identification of the reads that contain the reference allele

This file can be subsequently processed to extract a summary of counts of UMIs that support the REF and ALT alleles for every cell like so:

${CTAT_MUTATIONS_INSTALLDIR}/src/SingleCells/variant_cell_UMI_count_report.py \
     --vcf_sc_reads_tsv ${sample_name}.annot_pass_reads.vcf.sc_reads.gz \
     --output ${sample_name}.annot_pass_reads.vcf.sc_reads.variant_to_cell

chr_pos_variant cell_barcode    num_reads_w_variant     num_ref_matching_reads
chr10:100000235:C:T     ACGAGCCCACTATCTT        1.0     0.0
chr10:100000235:C:T     ATCTACTGTCTGATCA        1.0     0.0
chr10:100000235:C:T     CAAGTTGTCACCACCT        1.0     0.0
chr10:100000235:C:T     CATTCGCCACTTACGA        1.0     0.0
chr10:100000235:C:T     CCATGTCTCCTAGTGA        1.0     0.0
chr10:100000235:C:T     GCAGCCACATCGACGC        1.0     0.0
chr10:100000235:C:T     GCAGTTATCCCTCTTT        1.0     0.0
chr10:100000235:C:T     GGCGTGTCACGGACAA        1.0     0.0
chr10:100000235:C:T     GTACGTAAGAGGTACC        1.0     0.0
chr10:100000235:C:T     GTGCATATCCAGAAGG        1.0     0.0
chr10:100000235:C:T     GTTCGGGTCGGTGTTA        1.0     0.0
chr10:100000235:C:T     TCACGAATCAATAAGG        1.0     0.0
chr10:100000235:C:T     TGAAAGATCAGGATCT        1.0     0.0
chr10:100000235:C:T     TGACTAGTCATTCACT        1.0     0.0
chr10:100000235:C:T     TGAGCCGAGTTGTAGA        1.0     0.0

From this output, you can extract for every variant the cells that express the REF and/or the ALT allele, and explore these further using single cell analysis techniques - such as painting variant containing cells in a UMAP plot, or identifying those variants that show up in tumor cells as opposed to normal cells and likely represent somatic mutations:

Clone this wiki locally