How to align fastq files against a reference assembly

Question

[Priyanka]

Hello,

I am trying to align fastq files against the bacterial MRSA ATCC 33591 reference genome. The problem I am facing is that I have the reference assembly in fasta format with multiple sequences and upon creating index with hisat2-build I am not getting a good alignment rate. In fact most of the reads go unaligned.

I am curious to know if hisat2 can be used with genome assemblies directly or it needs to be converted to a single fasta file? Or am I doing something wrong.

I am new with bacterial genome and any advice on what to do or which tools to use will be helpful.

Thank you.

[hafiz.talhamalik]

hisat2 do take multi-fasta file..

[Priyanka]

So there is no problem with using the genome assembly as a reference with hisat2 right?

[samuel.a.odonnell]

Your genome assembly/reference genome should be within a single fasta file. You can easily join it together with cat. Does it work if you join it all?

[Priyanka]

My genome assembly is in one single fasta file with different headers for each contig and I am getting the result. However, the alignment rate is very low.

9557190 reads; of these:
  9557190 (100.00%) were paired; of these:
    9372275 (98.07%) aligned concordantly 0 times
    59908 (0.63%) aligned concordantly exactly 1 time
    125007 (1.31%) aligned concordantly >1 times
    ----
    9372275 pairs aligned concordantly 0 times; of these:
      19 (0.00%) aligned discordantly 1 time
    ----
    9372256 pairs aligned 0 times concordantly or discordantly; of these:
      18744512 mates make up the pairs; of these:
        18662286 (99.56%) aligned 0 times
        33977 (0.18%) aligned exactly 1 time
        48249 (0.26%) aligned >1 times
2.37% overall alignment rate

[hafiz.talhamalik]

check if alignment (whatever the rate is) covers the whole genome. ?? view bam files using IGV or Tablet. also check if you are using the same reference

[Priyanka]

Here is the bam file for poor alignment rate. There are regions where reads are getting mapped in different contigs. However, the alignment rate is very low. I am not sure the reason for this. I will try using some different aligner as well to see if alignment rate varies.

[Priyanka]

Here is the bam file for poor alignment rate. There are regions where reads are getting mapped in different contigs. However, the alignment rate is very low. I am not sure the reason for this. I will try using some different aligner as well to see if alignment rate varies.

[samuel.a.odonnell]

A few things I might look at: \ What happens if you blast one of your contigs in NCBI? \ And if you blast a few of your reads (the unmapped ones perhaps)? \ What are the regions where reads are aligning? Does it explain this very sparse alignment pattern? Can you align the same reads that were used to generate the assembly?

[Priyanka]

Thank you for the suggestions Samuel.

I did BLAST the reads and they were giving me hits for other bacteria, so I think the information about the reference given to me was perhaps not correct.

I am planning to map the reads to all the other bacterial genomes to see which genus my samples are mapping to the most.

[hafiz.talhamalik]

take a few hundred random reads from your sample file and try blast on those.