pileup output format

[charlie_sequencing]

Hey guys, I went through pileup format specifications on the website many times, but still have some silly questions. Would you please give me some hints? Thanks!

One sentence in http://samtools.sourceforge.net/pileup.shtml says that "The ASCII of the character following `^' minus 33 gives the mapping quality." Take the first line as an example:

seq1 272 T 24 ,.$.....,,.,.,...,,,.,..^+. <<<+;<<<<<<<<<<<=<;<;7<&

This line shows that there are 24 reads covering this chromosome position on the reference, and the first base (^+.) in the last read matches the reference base at the same position with a "mapping quality" of 10 (43 - 33 = 10).

I have two questions:

1. Does "mapping quality" in the sentence "The ASCII of the character following `^' minus 33 gives the mapping quality" mean "mapping quality" of that read base, or "mapping quality" of the read containing that base?

2. Why this "mapping quality" is only available for the first read base (starting with ^), but not for all the other 23 read bases at the same position?


Sorry for the stupid questions because I really can't figure out those details.

Thanks in advance for any insights you will provide!

[swbarnes2]

Single bases don't have mapping quality; how could they? Mapping quality is an attribute of a read, not a base. Look at the entries in the sam file for the reads crossing that base, each letter has a quality, but the read as a whole has a mapping quality.

[charlie_sequencing]

Single bases don't have mapping quality; how could they? Mapping quality is an attribute of a read, not a base. Look at the entries in the sam file for the reads crossing that base, each letter has a quality, but the read as a whole has a mapping quality.

Hi swbarnes2, thank you very much for your reply! So why the mapping quality is only available for one read, but not for the remaining 23 reads? I still have no idea about this. Do you have any clue?

[sdvie]

As far as I understand, the pileup only contains the bases read in all reads covering a given position (in your example: ,.$.....,,.,.,...,,,.,..^+.) and their base qualities (<<<+;<<<<<<<<<<<=<;<;7<&).

Then there are sort of special characters, that may appear in some lines of the pileup, but not necessarily:
The $. that indicates that this base is the last base on a read and
The ^+. that indicates that this base is the first base on a read. In the latter case, for some reason, the mapping quality of this read is given.

I am still not entirely clear on this either.

cheers,
Sophia

[charlie_sequencing]

As far as I understand, the pileup only contains the bases read in all reads covering a given position (in your example: ,.$.....,,.,.,...,,,.,..^+.) and their base qualities (<<<+;<<<<<<<<<<<=<;<;7<&).

Then there are sort of special characters, that may appear in some lines of the pileup, but not necessarily:
The $. that indicates that this base is the last base on a read and
The ^+. that indicates that this base is the first base on a read. In the latter case, for some reason, the mapping quality of this read is given.

I am still not entirely clear on this either.

cheers,
Sophia

Thank you, Sophia! I did see a lot of .$ in my pileup output, and will check if either .$ or $. (as you mentioned) indicates that it's the last base in the read matching the reference on the forward strand.

[swbarnes2]

Hi swbarnes2, thank you very much for your reply! So why the mapping quality is only available for one read, but not for the remaining 23 reads? I still have no idea about this. Do you have any clue?

The mapping quality is available for every single read in the pileup. It's placed at the first base of every read, wherever that falls. It would take up way too much space to put it at every single position of the read, for every single read. And how else could you expect to see all 23 of them at position 272 or wherever?

It's also available in the sam file.