MAQ error

[jyli]

I am using MAQ 0.7.1 and I recently encountered a weird errors:

-- maq-0.7.1
[ma_load_reads] loading reads...
[ma_load_reads] set length of the first read as 35.
[ma_load_reads] 17547636*2 reads loaded.
[ma_longread2read] encoding reads... maq: read.cc:106: match_info_t* ma_longread2read(const longreads_t*): Assertion `matches' failed.

I double checked that both query data and database data were OK. Any insight is appreciated.

[johnwhitaker]

I also got this error. Did you ever work out what was going wrong?

[gcarbajosa]

maybe you are loading too many reads at the same time?

from the maq's user manual:

"The best way to run this command is to provide about 1 to 3 million reads as input. More reads consume more memory"

I suggest that you split the file in smaller sets of 1 to 3 million reads as suggested. I've done it and it works faster and safer.

http://maq.sourceforge.net/maq-manpage.shtml

[johnwhitaker]

I got around the problem by increasing the memory to 16 Gigs.

[jyli]

So, it is memory usage problem then? Interesting though as the developer claimed that we do NOT need to split file, which I always did in the past.

[erichpowell]

I got around the problem by increasing the memory to 16 Gigs.

How did you increase the memory? I don't see any parameters to modify that?

I am very interested in this solution because I have encountered this error before and in past times I was able to get past this error by reducing the number of reads I was aligning, say, from 2M to 1M. In this case, I have dropped the number of reads as low as 750,000 and still the error persists.

[johnwhitaker]

I think I just specified that the job should be allowed more memory when submitting to our cluster.

[Brajbio]

I too got the similar error.

But surprisingly when i am feeding the single read input file (.bfq) of size 2.9 GB of each read 75 bp, its running sucessfully with MAQ but when i am giving the input file (.bfq) of size 3.0 GB (each read 75 bp), i am getting the error:

-- maq-0.7.1
[ma_load_reads] loading reads...
[ma_load_reads] set length of the first read as 75.
[ma_load_reads] 82785673*2 reads loaded.
[ma_longread2read] encoding reads... maq: read.cc:106: match_info_t* ma_longread2read(const longreads_t*): Assertion `matches' failed.
Aborted

I am pretty sure there is no memory problem , as i am using 64 GB RAM

I will greatly appreciate if someone can help me out to tackle this.

[Brajbio]

To my previous post i would like to add that......i had already successfully match this big size reads data once before with MAQ and even i have got the output as .map files, so i am pretty sure, size of reads and the memory is not the constrain to be pointed out. Still i am running the Maq with input of 2.9 GB of reads files (.bfq) having length read 75 bp each, which is processing perfectly.

[qiuyongquan]

Phred-like quality value

What does the following sentence mean?
"Over all the raw reads, the average Phred-like quality value at base 80 was Q26.6 and the average base position at which this declined below Q30 was 69.7."

How can I use maq to get Phred-like quality value?