fragments file generation via Sinto from CellRanger output

Question

[ntsopoul]

Hi,

I am following the instructions for the PASTA package (https://satijalab.org/seurat/articles/pasta_vignette.html). This package uses scRNA-seq data to infer alternative polyadenylation usage from scRNAseq data. It requires among many input files also a fragment file. The authors state the following must be done to obtain this file:

A fragment file (produced from the aligned BAM file, we recommend using the blocks function in sinto.

So I successfully installed Sinto but when I followed the instructions on the Sinto website I got a file with 0 bytes. As I understood you ought to use the possorted_genome_bam.bam file from CellRanger output.

this is my prompt:

sinto fragments -b possorted_genome_bam.bam  -f my.bed --barcode_regex "[^:]*" -p 4

and this is the output:

Function run_fragments called with the following arguments:

bam possorted_genome_bam.bam
fragments   my.bed
min_mapq    30
nproc   4
barcodetag  CB
cells   None
barcode_regex   [^:]*
use_chrom   (?i)^chr
max_distance    5000
min_distance    10
chunksize   500000
shift_plus  4
shift_minus -5
collapse_within False
func    <function run_fragments at 0x7fe9a85dfa60>

Function completed in  0.0 m 0.59 s

I also ran it without specifying barcode_regex but I got the same result.

In case you wonder what the possorted_genome_bam looks like: this is the output for samtools view possorted_genome_bam.bam | head -n 3

SRR17534016.9003356 16 chr1 3014927 0 63M30S 0 0 CCGACTAGGCCATCTTTTGATACATATGCAGCTAGAGACAAGAGCTCCGGGGTACTAGTTAGTCCCATGTACTCTGCGTTGATACCACTGCTT CCCCCCCCCC;C-CCCCCC-CCCCCCCCC-CC;CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC NH:i:10 HI:i:1 AS:i:62 nM:i:0 ts:i:30 RG:Z:young2F:0:1:unknow_flowcell:0 RE:A:I xf:i:0 CR:Z:CCGTTTATCGCTGTTC CY:Z:CC-CCCCCCCCCCCCC CB:Z:CCGTTCATCGCTGTTC-1 UR:Z:TTCAAGGTTCCA UY:Z:CCCCCCCCCCCC UB:Z:TTCAAGGTTCCA SRR17534016.11370570 16 chr1 3018673 1 92M 0 0 TTTGTTTTAGGATAAAATGTTCTGTAGATATCTGTCAAGTCCATTTGTTTCATCACTTCTGTTAGTTTCACTGTGTCCCTGTTTAGTTTCTG CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC NH:i:3 HI:i:1 AS:i:90 nM:i:0 RG:Z:young2F:0:1:unknow_flowcell:0 RE:A:I xf:i:0 CR:Z:CCGTAGGGTAGGATAT CY:Z:CCCCCCCCCCCCCCCC CB:Z:CCGTAGGGTAGGATAT-1 UR:Z:TGAATGGCTTCT UY:Z:CCCCCCCCCCCC UB:Z:TGAATGGCTTCT SRR17534018.9851676 16 chr1 3018686 1 63M30S 0 0 AAAATGTTCTGTAGATATCTGTCAAGTCCATTTGTTTCATCACTTCTGTTAGTTTCACTGTGTCCCATGTACTCTGCGTTGATACCACTGCTT CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC NH:i:3 HI:i:1 AS:i:62 nM:i:0 ts:i:30 RG:Z:young2F:0:1:unknow_flowcell:0 RE:A:I xf:i:0 CR:Z:CACGAATAGACCAGCA CY:Z:CCCCCCCCCCCCCCCC CB:Z:CACGAATAGACCAGCA-1 UR:Z:AGCTCCCGGGAT UY:Z:CCCC;CCCCCCC UB:Z:AGCTCCCGGGAT (biotools2) [nt793@eris2n4 outs]$ samtools view possorted_genome_bam.bam | head -n 1 SRR17534016.9003356 16 chr1 3014927 0 63M30S 0 0 CCGACTAGGCCATCTTTTGATACATATGCAGCTAGAGACAAGAGCTCCGGGGTACTAGTTAGTCCCATGTACTCTGCGTTGATACCACTGCTT CCCCCCCCCC;C-CCCCCC-CCCCCCCCC-CC;CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC NH:i:10 HI:i:1 AS:i:62 nM:i:0 ts:i:30 RG:Z:young2F:0:1:unknow_flowcell:0 RE:A:I xf:i:0 CR:Z:CCGTTTATCGCTGTTC CY:Z:CC-CCCCCCCCCCCCC CB:Z:CCGTTCATCGCTGTTC-1 UR:Z:TTCAAGGTTCCA UY:Z:CCCCCCCCCCCC UB:Z:TTCAAGGTTCCA

I understand that Sinto was written for scATAC-seq and I guess some more steps must follow in order to prepare it for Sinto. I am thankful for any help I can get.

[biofalconch]

What sort of aligner are you using? My guess is that the MAPQ in sinto is set too high :(

[ntsopoul]

Hi, I am using the output from celltanger which itself uses STAR. I just want to emphasize that I am dealing with RNA-seq and not atac seq.

[ntsopoul]

I ran

sinto fragments -b possorted_genome_bam.bam  -f my.bed --barcode_regex "[^:]*" -p 4 -m 0

the output is:

Function run_fragments called with the following arguments:

bam possorted_genome_bam.bam
fragments   my.bed
min_mapq    0
nproc   4
barcodetag  CB
cells   None
barcode_regex   [^:]*
use_chrom   (?i)^chr
max_distance    5000
min_distance    10
chunksize   500000
shift_plus  4
shift_minus -5
collapse_within False
func    <function run_fragments at 0x7fa1c01e7a60>

Function completed in  2.0 m 6.64 s

and the file is again 0 bytes. At least it was running longer now :P

[ntsopoul]

I also tried with: --shift_plus 0 --shift_minus 0 but the result is the same

(base) [nt793@dn018 outs]$ sinto fragments -b possorted_genome_bam.bam  -f my.bed --barcode_regex "[^:]*" -p 4 -m 0 --shift_plus 0 --shift_minus 0
Function run_fragments called with the following arguments:

bam possorted_genome_bam.bam
fragments   my.bed
min_mapq    0
nproc   4
barcodetag  CB
cells   None
barcode_regex   [^:]*
use_chrom   (?i)^chr
max_distance    5000
min_distance    10
chunksize   500000
shift_plus  0
shift_minus 0
collapse_within False
func    <function run_fragments at 0x2b3125ad53a0>

[biofalconch]

Hi, sorry for taking long to respond.

Since it is RNA (and I assume 3' based at that) sinto wont work. sinto requires paired end reads in order to work (usually one of the reads in RNA-seq is used for DNA storage ). Do you really need a fragment file for RNA-seq?

[GenoMax]

Satija lab is recommending that sinto be used in the link provided by OP in original post though.

What they say is use blocks function in sinto which I am not sure how it relates to sinto fragments command.

[ntsopoul]

The Satija lab just commented on github.

I will try their code and see if it works. In any case I will let you know in this thread.