[matlock bamfilt] --over filtered?

[zhaotao1987]

Hello, I did reads-mapping and sorting using the following, then I used matlock for the filtering.

## 1. make index file for the assembly.
        bwa index -a bwtsw $assembly
## 2. align hi-c reads
        bwa mem -t $cpu -5SP  $assembly $hic_1 $hic_2 |samblaster | samtools view -@10 -S -h -b -F 2316 > hic_reads.aligned.bam
## 3. sort reads
        samtools sort -@10 -n hic_reads.aligned.bam -o hic_reads.aligned.sorted.bam

filtering:

matlock bamfilt \
-i hic_reads.aligned.sorted.bam \
-o hic_reads.aligned.sorted.cleaned.bam

Then my raw 88G bam now reduced to ~0.9G. I'm very happy at first, also using hic_qc.py it seems the percentages are getting better.
But when I used such a result for 3ddna or allhic, it seems didn't work well probably due to very low signal. what would be the reason, could you give some advice, thank you very much!