SNP calling - is there an accepted Phred quality threshold?

[Francesco Lescai]

I just would like to know if there is an established Phred quality threshold in the literature in order to consider a SNP call reliable.
Provided the depth is higher than 30x (as showed in many papers), is the Phred Quality threshold to be chosen related to the coverage?
Are there commonly accepted criteria?

[apratap]

Hi

I am also looking at the thresholds that I could use to filter the SNP calls. I have seen people using SNP quality cut off at 50 but not in some paper that I could refer you to.

If anyone can educate me on what good thresholds could be it would help.

The parameters in question are :

1. SNP Quality
2. Depth
3. Avg Base quality of all bases taking part in the SNP call.

Thanks!
-Abhi

[nilshomer]

Hi

I am also looking at the thresholds that I could use to filter the SNP calls. I have seen people using SNP quality cut off at 50 but not in some paper that I could refer you to.

If anyone can educate me on what good thresholds could be it would help.

The parameters in question are :

1. SNP Quality
2. Depth
3. Avg Base quality of all bases taking part in the SNP call.

Thanks!
-Abhi

Given you have sufficient coverage, making you sure you see the variant allele on both strands at least once is a great filter.

[bioinfosm]

I have seen filtering out adjoining SNP calls is also useful. Many false positives are positions next to each other on reference.