Fastq Splitter vs Fastq de-lancer on Galaxy

Question

[pbm11]

Hello Everyone

I have 1 RNAseq file (.fastq format) which has both forward and reverse reads. How should I proceed for Hisat2 alignment? Should I use Fastq Splitter or Fastq de-lancer on this file to map it? If it so why?

Thanks in advance

[GenoMax]

Because HISAT2 does not accept interleaved reads you will have to separate read 1 and 2 into separate files.

[pbm11]

I just tried to align this file with Hisat2 (using library option -Pairend data from single interleved dataset) and this is alignment summary, what do u think about this?

42245991 reads; of these:
  42245991 (100.00%) were paired; of these:
    4086755 (9.67%) aligned concordantly 0 times
    34949638 (82.73%) aligned concordantly exactly 1 time
    3209598 (7.60%) aligned concordantly >1 times
    ----
    4086755 pairs aligned concordantly 0 times; of these:
      438370 (10.73%) aligned discordantly 1 time
    ----
    3648385 pairs aligned 0 times concordantly or discordantly; of these:
      7296770 mates make up the pairs; of these:
        4205058 (57.63%) aligned 0 times
        2340862 (32.08%) aligned exactly 1 time
        750850 (10.29%) aligned >1 times
95.02% overall alignment rate

[GenoMax]

Looks good. Your data appears to be aligning well.

[pbm11]

Thanks!!! Incase I am not using this option and need to split the file into 2 parts, which tool should I use: Splitter or delancer?

[GenoMax]

Hmm. You did not de-interleave your reads and HISAT2 accepted that file?

Looks like Galaxy may be doing something under the covers to split the files and then feed them to HISAT2 or HISAT2 accepts interleaved reads (though the manual seems to make no mention of it). It appears to have worked right.

I looked up fastq splitter and de-interlacer on Galaxy. Looks like only difference is that splitter expects reads to have identical lengths. But otherwise you could use either. Or so it seems.

I personally use reformat.sh from BBMap suite to de-interleave reads.