I am working on WGS, and I have generated two pairs-end reads (total 4, two forward, and 2 reverse)(the difference is in their DNA purification method). Now, I want to combine these two pairs of reads to form one pair-end reads (1 forward, and 1 reverse), so that there is some kind of complementarity of information. Let me know how to do, or if simple copying one file to others will work?
I think you are very lucky if the recommended solution using simple concatenation of the gz’d readsets has worked in your processing pipeline. ’cat’ is ignorant of the binary file structure being concatenated so with the solution adopted you have ended up with two independent compressed datastreams (readsets) one immediately following the other. Many bioinformatics applications will be expecting a gz file to contain a single read set and will ignore any data streams subsequent to the first encountered in a gz readset file. You may also encounter system memory errors as it is common for applications to do an initial seek to end of file and read the last 4 bytes which in the gz file format contains the uncompressed datastream size mod 2^32. This datastream size will be the memory allocation size used which Is that of the uncompressed final datastream rather than that of all data streams Combined.
have a look at gzcat and related applications as to how to combine multiple compressed data streams into a single compressed datastream which downstream bioinformatics applications can correctly process as if processing a single sequencer readset.
Stuart Stephen Exactly, I was wondering about the same too. How can someone recommend to use cat for compressed file. And irony is, it is also getting recommended. Its funny to see the confident trade-off of wrong information and appreciation..
Abhijeet Singh, I actually used cat with the extracted file. Please think of other possibilities before replying like that. Not cool.
Adarsh Singh accepted solution showing ‘cat’ used to concatenate two gz’d files. If different solution was used - ’cat’ on the extracted fasta/fastq files - then this new solution should be the published answer so as to prevent ambiguity. Answers accepted as being the best should reflect the adopted answer by the questioner so others with the same question may be guided by current best practice. I agree with the sentiments expressed by Abhijeet Singh as they were expressed in the context of the currently accepted best answer - this answer being incorrect and not best practice.
I will say no more in response to the specific question originally raised.
- Badges
- Science topic
- Similar topics
- Filing