Dear Concern,
I'm a relatively new MS-GF+ user, and have been trying to understand the significance of the different statistical measures returned in the output. I initially chose Qvalue = 0.01 as a cutoff, but then I noticed lots of PSMs meeting this criteria have a Evalue higher than 0.01. The documentation suggests the Evalue should be low ("the lower the better"), but is there any recommendation about whether a Evalue-based cutoff should also be incorporated to discount false positives?
On a related note, I ran MS-GF+ and MS-GF+percolator on the same samples and with otherwise identical parameters, but if I just filter both outputs by PSM Qvalues - vanilla MS-GF+ consistently returns a higher number of PSMs and peptides compared to MS-GF+percolator, which I understand is contrary to expectations. This leads me to wonder whether the E-value cutoff is also important in evaluating MS-GF+ results.
Thanks again and very best regards,
Hassan
Dear Concern,
I'm a relatively new MS-GF+ user, and have been trying to understand the significance of the different statistical measures returned in the output. I initially chose Qvalue = 0.01 as a cutoff, but then I noticed lots of PSMs meeting this criteria have a Evalue higher than 0.01. The documentation suggests the Evalue should be low ("the lower the better"), but is there any recommendation about whether a Evalue-based cutoff should also be incorporated to discount false positives?
On a related note, I ran MS-GF+ and MS-GF+percolator on the same samples and with otherwise identical parameters, but if I just filter both outputs by PSM Qvalues - vanilla MS-GF+ consistently returns a higher number of PSMs and peptides compared to MS-GF+percolator, which I understand is contrary to expectations. This leads me to wonder whether the E-value cutoff is also important in evaluating MS-GF+ results.
Thanks again and very best regards,
Hassan