1
$\begingroup$

I performed a blastn search of NGS data against ssRNA database download from Internet, with a expected value 10-4. The size of NGS data reads is of 125 bp.

I have analyzed the blast results of the libraries. These libraries are from different habitats. The most difficult habitat, oakwood, produced a blast results with a maximum length of 51 bp. These results have a qcov of 100 (in most of the cases).

I would like to know if these low length sizes comes from degradation?

Should do I filter these samples with a length = 125 bp with the aim to obtain results with better quality or with a qcov of 100 is enought to get reliable results?

Thanks you in advance.

$\endgroup$
7
  • 1
    $\begingroup$ I think you should maybe rely on the hitscore, because it integrates both overlap length and % of identity. In addition, you may have a look at MMSeqs2 as an alternative to BLAST. It is much faster, and has tutorials to blast ssRNA db such as SILVA (github.com/soedinglab/mmseqs2/wiki). $\endgroup$ May 15 '20 at 10:07
  • $\begingroup$ What would "better results" mean? What is wrong with what you get now? If your queries have a maximum length of 125, why would you search for >= 125? How can you have a hit with a length greater than your query sequence? Please edit your question and clarify. $\endgroup$
    – terdon
    May 15 '20 at 12:37
  • 1
    $\begingroup$ Could you please explain your RNA extraction protocol from field collection to lab? $\endgroup$
    – M__
    Jun 29 '20 at 19:23
  • 1
    $\begingroup$ The RNA extraction was extracted with a CTAB kit from tissue plant of different plants species. This plant species comes from the habitat wilder. My concern is that the blast results gave to me hits with a high qcov and pident and low length. $\endgroup$ Jun 29 '20 at 19:50
  • $\begingroup$ could please give more information? which ssRNA database are you using? and how your reads have been preprocessed (trimmed)? $\endgroup$ Jun 30 '20 at 4:25
1
+100
$\begingroup$

I had a think about this one and am stumped. I thought there could be a connection with acute oak disease (AOD) or else chronic oak disease (COD), which are emergent infections threatening the population. However, they appear to be fungal and bacterial infections rather than viral.

Single stranded RNA plant viruses are very common, more so than in mammals, but to achieve the results you are obtaining the trees would need to be really sick if the predominent RNA was viral. Short viral RNAs are produced regularly by viruses as defective RNA and its a major area of research, considering a mechanism of defective interference,

In context it would be weird if this RNA swamped the tree/plant, you need an actively replicating virus for this stuff to appear. So I don't think its either viral or defective viral RNA.

The antiviral response in mammals can involve RNase and this could explain your degradation. Essentially, mammals (probably vertebrates) kick-back hard when virus appears. However, I couldn't really find evidence of this in plants.

I don't think the results you obtained have a clear biological explanation in pathogenesis. The best explanation is it was wet lab-technique, e.g. library construction resulted in overdigestion/fragmentation. The only thing I can suggest is template assembly using a normally distributed plant/oak RNA library.

I was probably overthinking in forging a link with an emergent (tree) pathogen.


enter image description here!

$\endgroup$
6
  • $\begingroup$ Thanks you Michael. Your answer looks fit better with my results. $\endgroup$ Jul 5 '20 at 11:36
  • $\begingroup$ I will ask to the sequencing company asking them about the library construction. Thank you again $\endgroup$ Jul 5 '20 at 11:40
  • $\begingroup$ Thanks you are too kind, but emergent viruses is what I do (albeit for humans not plants) and I couldn't get this one to make sense. Ask them for the Agilent library outputs, both good runs.and 50bp run. I've posted an example above. $\endgroup$
    – M__
    Jul 5 '20 at 12:29
  • $\begingroup$ Hi @AdriánP.L. please let us know how you get on here: one way or another we'll get this resolved if not directly via collaboration. $\endgroup$
    – M__
    Jul 7 '20 at 16:11
  • 1
    $\begingroup$ Dear @Michael I though It was solved, your answer looks fitting well with the trouble . $\endgroup$ Jul 7 '20 at 17:05

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.