Different result by bowtie2

Question

[daffodil]

I ran the following Bowtie2 command, but my results show fewer uniquely mapped reads compared to my supervisor's results, who obtained 22,000,000 uniquely mapped reads. The command I used is:

bowtie2 -x /sw/data/iGenomes/Mus_musculus/UCSC/mm10/Sequence/Bowtie2Index/genome \
  -U cut_trimmed.fastq.gz \
  --no-unal \
  -X 700 \
  -S cut.sam

I have also used cutadapt for trimming of single read. What could be the reason for this discrepancy in the number of uniquely mapped reads?

[dthorbur]

There are many reasons you might get different number of reads with a tool like Bowtie2. The tool uses heuristics to be fast and efficient, so even two runs with most of the same parameters will differ.

How large is the difference between your two datasets? Are all the other parameters identical in your runs?

[biofalconch]

Just adding to this, it would be nice if your supervisor would use the --seed flag to obtain the same results.

[daffodil]

Thank you for your response. All of them are the same, however, he took approximately 22 million uniquely mapped reads.

[GenoMax]

If you need deterministic results (identical each time for each run) then you may have to use a program that allows for it. bbmap the aligner is one such program.

my results show fewer uniquely mapped reads

Define "fewer".