about bowtie parameter --fr and --rf

Question

[ch8316f5eyu]

I was using bowtie2 to analyze ATAC data. The raw data is paired. The result is wired that the percent of reads aligned concordantly is very low but the percent of reads aligned is very high. Then I changed parameter to "--rf", the percent of reads aligned concordantly was higher but the sequencing instrument is illumina. Does anyone know what is wrong?

--fr:

~/miniconda3/envs/bam/bin/bowtie2 -5 1 --very-fast -p 6  -x GRCh38_noalt_as --fr  -I 0 -X 2000 -1 ATAC_EPS_1_R1.fq.gz -2 ATAC_EPS_1_R2.fq.gz > xx
5222770 reads; of these:
  5222770 (100.00%) were paired; of these:
    3954657 (75.72%) aligned concordantly 0 times
    1058713 (20.27%) aligned concordantly exactly 1 time
    209400 (4.01%) aligned concordantly >1 times
    ----
    3954657 pairs aligned concordantly 0 times; of these:
      2240623 (56.66%) aligned discordantly 1 time
    ----
    1714034 pairs aligned 0 times concordantly or discordantly; of these:
      3428068 mates make up the pairs; of these:
        1151320 (33.59%) aligned 0 times
        97212 (2.84%) aligned exactly 1 time
        2179536 (63.58%) aligned >1 times
88.98% overall alignment rate

--rf

~/miniconda3/envs/bam/bin/bowtie2 -5 1 --very-fast -p 6  -x ~/ref/Align_index/bowtie2/hg38/GRCh38_noalt_as --rf  -I 0 -X 2000 -1 /home/sys4/ATAC_hEPSC/trimed/ATAC_EPS_1_R1.fq.gz -2 /home/sys4/ATAC_hEPSC/trimed/ATAC_EPS_1_R2.fq.gz > xx
5222770 reads; of these:
  5222770 (100.00%) were paired; of these:
    1789000 (34.25%) aligned concordantly 0 times
    2249944 (43.08%) aligned concordantly exactly 1 time
    1183826 (22.67%) aligned concordantly >1 times
    ----
    1789000 pairs aligned concordantly 0 times; of these:
      982313 (54.91%) aligned discordantly 1 time
    ----
    806687 pairs aligned 0 times concordantly or discordantly; of these:
      1613374 mates make up the pairs; of these:
        1163445 (72.11%) aligned 0 times
        134697 (8.35%) aligned exactly 1 time
        315232 (19.54%) aligned >1 times
88.86% overall alignment rate

[ATpoint]

Is chrM in the reference genome?

[ch8316f5eyu]

Yes, and I remove reads mapped to chrM by samtools in later step. Do you think reads mapped to chrM is the reason? But why?

[ATpoint]

I thought missing chrM could be the reason for the high rate of unaligned reads. I do not know what is going on here. Did the sample look normal on the Bionalayzer? banding pattern visible?

[ch8316f5eyu]

This is the insertion size which is also abnormal, but I can't connect the two problem.

[ATpoint]

Most reads in the first peak are slightly below 100bp, and then there is the banding pattern, looks rather normal I guess. Try to blast some of the unmapped reads, maybe that gives a hint what is going on.