Macs2: predicted fragment length less than 2*tag length

Question

[langya]

I am using

macs2 2.1.1.20160309 to call peaks on narrow peak ChIP-Seq data set that is sequenced with Single Read 101 bp and then trimmed the reads with TRIMMOMATIC and mapped to bowtie2 (with good mappability), I then called Peaks with macs2 with following command: macs2 callpeak -t $1 -c $2 -f BAM -g hs -n $1 -B --SPMR -q 0.05

But then in the macs2 output, there is this warning:

2 predicted fragment length is 93 bps

2 alternative fragment length(s) may be 93,463,534 bps

2 Since the d (93) calculated from paired-peaks are smaller than 2*tag length, it may be influenced by unknown sequencing problem!

Could anyone help me out?

[ATpoint]

Could be a poor ChIP efficiency because of a bad antibody, but that is guessing. Can you post a screenshot of the IGV (or whatever browser) for a representative region, like a 10kb window or so, centering on what you would call a peak? How to add images on biostars