Chip-Seq's mapping rate problem.(too low)

Question

[leenaehyeon]

Chip-Seq's mapping rate problem. I've run FastQC and result is not bad but mapping rate is horrible (read tophat_out/align_summary.txt)

Left reads:
          Input     :  14775968
           Mapped   :    133948 ( 0.9% of input)
            of these:     36875 (27.5%) have multiple alignments (1612 have >20)
Right reads:
          Input     :  14775968
           Mapped   :    123596 ( 0.8% of input)
            of these:     30111 (24.4%) have multiple alignments (1430 have >20)
 0.9% overall read mapping rate.

Aligned pairs:     91752
     of these:     15121 (16.5%) have multiple alignments
                    6000 ( 6.5%) are discordant alignments
 0.6% concordant pair alignment rate.

First, using bbmap --> mapping(tophat2)

I don't know why.

[Asaf]

If it's Chip-Seq why did you use Tophat? Although I don't think this is the problem. Try to run FastQC after bbmap

[leenaehyeon]

I usually using tophat2 for RNA-seq, so I tried same way for Chip-seq.

Is there another way for Chip-seq analysis tool?

[jotan]

Bowtie/Bowtie2 is normally used for ChIP-seq.

http://bowtie-bio.sourceforge.net/index.shtml

It runs within Tophat so it will already be installed. It also comes with ready-made genome libraries. Maybe try aligning against some pre-made libraries as well just in case something went wrong with your genome build.

[jotan]

I'm really sorry for this question but I have to ask. Are you sure you're working on the right genome? I've accidentally tried to align mouse data to human genome and vice versa.

[leenaehyeon]

My boss said that genome download from UCSC.. :(((( ..hmm..

I'll try another way.. Thank you for reply!

[leenaehyeon]

#[1] bbmap
/home/Downloads/bbmap/bbduk.sh \
  in=A1.fastq \
  in2=A2.fastq \
  out=A1_bbduk.fastq \
  out2=A2_bbduk.fastq \
  ref=/home/Downloads/bbmap/resources/truseq.fa.gz \
  k=13 \
  ktrim=r \
  qtrim=t \
  trimq=10 \
  minlength=20

#[2] mapping
tophat2 \
  -o tophat_out_A \
  -G /home/Documents/genome/mm10/mm10.gtf \
  --transcriptome-index /home/Documents/genome/mm10/mm10_transcripts \
  --library-type fr-firststrand \
  /home/Documents/genome/mm10/mm10 \
  A1_bbduk.fastq A2_bbduk.fastq

Is that wrong?