#2 compatible with different assemblies

README.md

@@ -9,12 +9,14 @@ Developed by Richard Koche, adapted to BIH cluster by @haasek and maintained by
 - [Installation](#installation)
 - [Usage](#usage)
 - [Run circle-enrich-filter example](#run-circle-enrich-filter)
+- [Run circle-enrich-filter with PDX data](#run-pdx)
 - [Citation](#citation)
 - [License](#license)
 
 ## Installation <a name="installation"></a> 
 
-Prequisites are conda and pgltools v2.2.0. Compatible with `python 2.7` and `python 3.7`.
+Prequisites are conda and pgltools v2.2.0. Compatible with `python 2.7` and `python 3.7`.<br/>
+Compatible with human (`GRCh38.p13`, `GRCh37.p13`, `hs37d5`, `hg38`, `hg19`) and mouse (`GRCm38`) genome assemblies.
 
 ### 1. Create conda environment
 
@@ -90,7 +92,7 @@ Options:
 
 ## Run circle-enrich-filter example <a name="run-circle-enrich-filter"></a>
 
-Run test example using the `2:14508020-18508849` region from a CHP212 cellline sequenced using the Circle-seq Illumina sequencing.
+Run test example using the `2:14508020-18508849` region from CHP212 human cellline sequenced using the Circle-seq Illumina sequencing.
 
 ```
 bash run_CircleEnrichFilter.sh -i example/output/chp212_2_14508020_18508849.bam -o example/2_14508020_18508849.fa 
@@ -107,6 +109,11 @@ The pipeline generates the all files under `example/ouput`. The final enriched r
 | 5       | all counts (coverage all reads spanning the junction) | 
 
 
+## Run PDX samples with circle-enrich-filter <a name="run-pdx"></a>
+
+In case you run PDX samples, please consider annotating all chromosomes from the mouse using `m.*`, .e.g `m.1` or `m.chr1`. 
+
+
 ## Citation <a name="citation"></a>
 
 Koche, R.P., Rodriguez-Fos, E., Helmsauer, K. et al. Extrachromosomal circular DNA drives oncogenic genome remodeling in neuroblastoma. Nat Genet 52, 29-34 (2020). 

run_CircleEnrichFilter.sh

@@ -126,14 +126,14 @@ grep -v "^#" $peakfile | cut -f2-4,6 > $peakbed
 # Output merged regions of enrichment
 # Edges may not be perfect due to where enrichment block falls, but works in most cases and can be corrected below
 mergedbed=${peakfile/%.txt/.enriched.merged.orig.bed}
-grep -v "^#" $peakfile | cut -f2-4 | grep -v "random\|chrUn\|GL\|NC_\|hs37d5\|EGFP\|KI" | sort -k1,1 -k2,2n | bedtools merge -d $mergedist -i stdin > $mergedbed
+grep -v "^#" $peakfile | cut -f2-4 | grep -v "alt\|random\|chrU\|GL\|NC\|hs37d5\|EGFP\|K\|ML\|JH" | sort -k1,1 -k2,2n | bedtools merge -d $mergedist -i stdin > $mergedbed
 
 
 # Output merged reads
 # (sometimes more accurate delineation of circle junctions, but can fall into trap of cascading reads outside of enriched region)
 # First get bam to bed
 bam2bed=${inbamnodir/%.bam/.bam2bed.bed}
-samtools view --threads $nthreads -bq $qfilt $inbam | bedtools bamtobed -i stdin -splitD | cut -f1-3 | grep -v "random\|chrUn\|GL\|NC_\|hs37d5\|EGFP\|KI" | sort -k1,1 -k2,2n | bedtools merge -d 250 -i stdin > $bam2bed
+samtools view --threads $nthreads -bq $qfilt $inbam | bedtools bamtobed -i stdin -splitD | cut -f1-3 | grep -v "alt\|random\|chrU\|GL\|NC\|hs37d5\|EGFP\|K\|ML\|JH" | sort -k1,1 -k2,2n | bedtools merge -d 250 -i stdin > $bam2bed
 
 # Then get bam2bed overlap with enriched blocks, but only keep segments with >5 reads
 # (this is for edge fine-tuning, not circle calling)
@@ -322,8 +322,8 @@ bamCoverage --extendReads 0 --minMappingQuality $qfilt --ignoreDuplicates --binS
 ## First, chrM plots:
 # get chrM coords for this genome assembly, based on bam header (differs depending on used genome build):
 chrMcoord="chrM.bed"
-#samtools view -H $inbam | grep "SN:chrM" | awk -v OFS='\t' '{ split($3,a,":"); print "chrM",1,a[2]}' > $chrMcoord
-samtools view -H $inbam | grep -P "SN:(chr)*MT" | awk -v OFS='\t' '{split($3,a,":"); print "MT",1,a[2]}' > $chrMcoord
+samtools view -H $inbam | grep "SN:chrM" | awk -v OFS='\t' '{ split($3,a,":"); print "chrM",1,a[2]}' > $chrMcoord
+samtools view -H $inbam | grep "SN:MT" | awk -v OFS='\t' '{split($3,a,":"); print "MT",1,a[2]}' >> $chrMcoord
 
 # meta plots
 # create data matrix (one site, so here a simple vector)