Error! not matched in input files

[sroener]

Hi,

I am working on reprocessing some samples and I want to use NGmerge to properly merge the PE reads. For this I convert an existing .bam file to fastQ files and use them as input for NGmerge. I execute the program like this:

NGmerge -w resources/qual_profile.txt -u 41 -n 8 -z -1 FILE_R1.fastq.gz -2 FILE_R2.fastq.gz -o FILE_merged.fastq.gz -f FILE_nonmerged -l FILE.log

For most samples, everything works like a charm, but for some I get errors like this:
Error! @HISEQ_172:2:2211:1315:83788 BC:Z:NAGCGTTANGAGTCAA: not matched in input files

Any idea what the problem might be?

[jsh58]

This error occurs when the paired reads' headers do not match. As the README states:

The input files must list the reads in the same order. The program requires that the paired reads' headers match, at least up to the first space character (or whatever alternative character is specified by -t).

[sroener]

Thx for the fast reply and pointing out when the error occurs.

Apparently, I had the misconceptions that these reads would also be written to the file specified with the -f flag?

Interestingly, if I look at the _R1 and _R2 files and search for the read headers mentioned in the error message, I find identical headers in both files.

As the log files are showing what I would identify as normal behaviour of NGmerge and that it takes a while for the error to occur, I would assume that there are only a few reads with non matching headers. Any suggestion on how to further investigate this?

Edit:
Additional filtering for singletons reads solved the problem. Still confused why some of the error IDs are present in both R1 and R2 and can be found by grep.