Hey, i am tasked with assembling a metagenome. From my predecessor i got in fasta format one file that contains:
1) normal paired-end reads (forward and reverse), about 75 mio reads each, read length 20-135 bp
2) some pre-assembled reads, where forward and reverse read were assembled into 1 read. That one read is now tagged as forward, and the original forward and reverse reads, it was created from, are no longer present. Those make up 22 mio reads with average read length 200
3) about 0,5 single reads, the partner was filtered out during QC
Now i can split the fasta file into these 3 categories, but i don't know how to deal with these pre-assembled reads. I have never seen such pre-assembled reads i don't know how to deal with them. (They were apparently provided like this from the sequencing company) I wanted to use metaspades for the assembly, but it does not take single-end libraries. If i only use the paired-end reads, then i loose 22 mio high quality reads from my sample.
Is there a metagenome assembler that takes in both paired and single-end reads for an assembly, or does anyone have some ideas how to incorporate these pre-assembled reads into the assembly?
Many thanks in advance
Do reads from (1) contained in (2) or they are two different groups? You have quite a mess there as reads are usually in fastq format, you should probably start the assembly from scratch. There are good assembly pipelines out there that start with the raw fastq files.