Question

How to calculate my own phred quality score

0

Entering edit mode

8.9 years ago

clear.choi ▴ 30

I am planning to calculate my own phred quality score. I need to share some idea how to calculate my own Q Score. I have sequences like below:

AAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAA
                    ....
TAAAAAAAAAAAAAAAAAAAAA

So in the above case, There is A: 100 and T: 1 on the first base. So that position has 99% accuracy. And from 99% accuracy how I can get Q = 20?

What if there is A: 151 and 1 T on the first position? Anyone knows how to calculate that score from the percentage?

Thank you

phred-quality-score fastq Qscore • 12k views

ADD COMMENT • link updated 2.3 years ago by Ram 44k • written 8.9 years ago by clear.choi ▴ 30

Ram · Accepted Answer · 2015-12-22

From Wikipedia page:

$Q=-10\ \log _{10}P$

Keep in mind that the original Phred score is intrinsic to the sequencing machine, which determines the score based on peak shape and resolution at every base. Assuming you still want to represent your percentages as Phred scores, you have to use the probability of the wrong base in the above equation. For example, if a base is 99% of the times 'A', then the probability of the base not being 'A' is 0.01. And Q score can then be derived as

Q = -10 * log10(0.01) = 20

Cheers