1
$\begingroup$

What's the best way to test if a pair of Illumina FASTQ files from blood DNA extraction contain any reads from viral/bacterial infection from DNA in human blood? Thanks in advance.

EDIT:

Looking at the available options, I started testing Kraken for samples where I have seen an indication of viral reads, I will update with results / further questions.

$\endgroup$
1
  • $\begingroup$ There are some people voting to close this question as too broad. Perhaps you could indicate what are your requirements and what options do you consider $\endgroup$
    – llrs
    Apr 24 '19 at 13:57
2
$\begingroup$

If you're talking RNA, I have a blog post from a while back about Diamond/MEGAN:

http://achri.blogspot.com/2017/04/data-for-free-viral-metagenomics-in.html

In principle this will work for DNA if you have a non-trivial number of pathogen reads (since they tend to be gene dense in their genomes). Otherwise I use Kraken on a database for all the NCBI nt database, but this is slower and doesn't let you sanity check the false positive taxonomy assignments as well as I describe in the post. Of course, ideally, you run multiple methods if you have the resources to do so.

$\endgroup$
2
$\begingroup$

I think you could try pathoscope for this.

$\endgroup$
1
$\begingroup$

To know if there are bacterial genomes in the sequenced sample I have used an approach based on kmers analysis. An useful tool to do that is kmerfinder which uses a kmers database built from bacterial genomes. Quoting from the kmerfinder web site:

The method is used to find the best match (species identification) to the reads in one or more fastq files or one fasta file in a (kmer) database produced using the KMA program.

I haven't use it for viruses genome analysis, but conceptually it should be work as well. Hope this would be helpful.

$\endgroup$
1
$\begingroup$

You may find this blog post helpful: Identifying viral sequences in TCGA data using Kraken and Centrifuge.

Our goal was to evaluate the performance of Centrifuge and Kraken for detecting viral transcripts across a cancer cohort. Centrifuge and Kraken produce very similar results and consistently detect HPV strains in cervical cancer.

There is also a review Overview of Virus Metagenomic Classification Methods and Their Biological Applications, but they focus specifically on viruses:

We evaluated 49 published computational classification workflows for virus metagenomics in a literature review. To this end, we described the methods of existing workflows by breaking them up into five general steps and assessed their ease-of-use and validation experiments. Performance scores of previous benchmarks were summarized and correlations between methods and performance were investigated. We indicate the potential suitability of the different workflows for (1) time-constrained diagnostics, (2) surveillance and outbreak source tracing, (3) detection of remote homologies (discovery), and (4) biodiversity studies.

$\endgroup$
0
$\begingroup$

What type of NGS data is it? If smallRNA or RNAseq data, you can simply assemble the reads into longer contigs and do the blast search using your reference viral or bacterial sequences.

$\endgroup$

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.