Computing A Short Read'S Score

Question

[Raygozak]

Hi, if i have the following read:

 @HWI-ST261:7:1:16113:8289#0/1
.CTCGGATAACCGTAGTAATTCTAGAGCTAATACGTGCAACAAACCCCGACTTCCGGGAGGGGCGCATTTATTAG
+
BWURTY[YYVacaccccccca_ccc\cc_ccccccccac_c_aYVcccc_c_accc_ZccUUUUVYV_c_[ZVV^

Where the 4th row is the score. as far as i understand each base has a probability which is interpreted as the quality, how can I compute the overall sequence read quality and be able to say the quality of the read is >50% for example. i know there several types of scores.

Thanks

Answer 1

[brentp]

With just the sequence, you can calculate the number of Q30 or Q20 bases -- the number of bases with scores greater than 30 and 20 respectively. So you could say X% of the bases have a quality score greater than 20. Once you have done the mapping, then you'll also get a mapping quality which is a single value that indicates the quality of the mapping.

Answer 2

[JC]

The overall quality score can be calculated with the average of the individual base scores (in this case is Phred+64 scale), but this average only represents how "good" is your read and can be biased if the Illumina's pipeline used "chastity" (you will see a run of B's close to the end).

To understand the scale, please check: http://en.wikipedia.org/wiki/Phred_quality_score and http://en.wikipedia.org/wiki/FASTQ_format