I wrote a "script" to pull out reads from a huge fastq file in an iterative manner, by finding homology to the previous sequence.

It should be relatively easy to overlap them and assemble the gene, however traditional assemblers like SPAdes fail due to obviously low coverage.

I am hesitant to just feed it more faked reads, when I know I have all the data necessary to just overlap them into a single sequence.

The reads are set up like this:


What kind of assembler is out there where I can feed it this kind of terribly-formatted data?


Miniasm will probably work for this; it's designed to be an ultra-fast, greedy assembler, which has been used for people doing real-time assemblies from MinION data. Because of its greedy nature, it can misassemble regions where the overlapping fragments are ambiguous, so it's worth using some other method in addition to miniasm to confirm that the assembly is correct.

By default it looks for overlap regions that have a coverage of 3 reads when joining (option -c), but I assume that could be reduced down to 2 reads, as in your example (which has a minimum overlapping coverage of 2).

  • 1
    $\begingroup$ Thanks for the suggestion. This actually wound up working in a roundabout way! $\endgroup$
    – Ryan Ward
    Oct 19 '20 at 4:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.