Skip to content

Latest commit

 

History

History
34 lines (18 loc) · 1.33 KB

preprocessing.md

File metadata and controls

34 lines (18 loc) · 1.33 KB

Preprocessing

Read alignment and peak calling

We used Kundaje pipeline for aligning pair ended reads to hg19 and removing duplicates as scABC.

bds_scr [SCR_NAME] atac.bds -align -species hg19 -species_file [SPECIES_FILE_PATH] -nth [NUM_THREADS] -fastq1_1 [READ_PAIR1] -fastq1_2 [READ_PAIR2]

The resulting bam files were merged using samtools.

samtools merge [AGGREGATE_BAM] *.trim.PE2SE.nodup.bam

The merged bam file was then used as input into MACS2 to call merged peaks for later analysis.

bds_scr [SCR_NAME] atac.bds -species hg19 -species_file [SPECIES_FILE_PATH] -nth [NUM_THREADS] -se -filt_bam [AGGREGATE_BAM]

Preprocessing

scATAC-seq data is required in peak count matrix as inputs. We can import scATAC-seq preprocessing function in scale module to preprocessing as scABC

from scale.utils import sample_filter, peak_filter, cell_filter

We filtered peaks that presented in >= 10 cells with >= 2 reads by peak_filter function

data = peak_filter(data)

We filtered cells that >= number of peaks / 50 by cell_filter function

data = cell_filter(data)

Or combine peak_filter and cell_filter

data = sample_filter(data)