Screen.seqs result varying

I have a data set of 2x150 reads of 54 pairs of 16S v4 metagenomic sequences from NCBI sra of gastritis patients. When I previously ran the sequences through mothur, the screen.seqs after silva alignment removed sufficient number of sequences.

mothur > screen.seqs(fasta=current, count=current, start=2, end=13426)
Using Ulcer_Donors\stability.trim.contigs.count_table as input file for the count parameter.
Using Ulcer_Donors\stability.trim.contigs.unique.align as input file for the fasta parameter.

Using 4 processors.

It took 1278 secs to screen 3544746 sequences, removed 3505876.

Running command: remove.seqs(accnos=Ulcer_Donors\stability.trim.contigs.unique.bad.accnos.temp, count=Ulcer_Donors\stability.trim.contigs.count_table)
Removed 5370441 sequences from Ulcer_Donors\stability.trim.contigs.count_table.

Output File Names:


Output File Names:

It took 2067 secs to screen 3544746 sequences.

When I am running the sequences again with the same parameters and commands the result is different. Why is it so?
mothur > screen.seqs(fasta=current, count=current, start=1, end=13424)
Using Ulcer_Donors/stability.trim.contigs.good.count_table as input file for the count parameter.
Using Ulcer_Donors/stability.trim.contigs.good.unique.align as input file for the fasta parameter.

Using 64 processors.

It took 52 secs to screen 2417753 sequences, removed 15177.

Running command: remove.seqs(accnos=Ulcer_Donors/stability.trim.contigs.good.unique.bad.accnos.temp, count=Ulcer_Donors/stability.trim.contigs.good.count_table)
Removed 18156 sequences from Ulcer_Donors/stability.trim.contigs.good.count_table.

Output File Names:


Output File Names:

It took 72 secs to screen 2417753 sequences.

This time I had put in an extra screen.seqs (to put in minlength ) which reduced the number of sequences sufficiently after make.contigs (to put in maxee, deltaq, insert). Could it be because of this?

Hi there - it looks like your start and end positions in the screen.seqs command are different in the two runs. That’s likely the reason.