0
$\begingroup$

This is my first time in this area of research. I am working on 95 varieties of bambara groundnut. I have done the agromorphological characterization of these varieties as well as the genetic diversity using DArT markers. Based on the charaterization result, i have selected 10 varieties to be subjected to water stress with the objective of identifying genes that are responsible for tolerance in this plant. My challenge now is: 1. bambara groundnut does not have a reference genome, so will i be able to use GWAS in getting these genes from the result of the DArTSeq. 2. Bearing in mind that the whole 95 varieties characterized were not subjected to water stress, do i need to set up another experiment entirely for the water stress and go ahead to make use of RNASeq to get the genes? 3. It is obvious that some genes will be upregulated and some will be downregulated, how do i identify these genes and how do i know if they are expressed at the onset of stress, during, or after stress? 4. when is the appropriate time to take my samples, i was considering taking at 50% flowering stage. Is this good? Any help and insightful articles will be greatly appreciated.

$\endgroup$
  • $\begingroup$ The choice of the experiment to do depends on the question you are trying to ask. RNAseq will give you a direct readout of gene expression which you can relate to a particular condition. GWAS is more difficult to interpret. $\endgroup$ – WYSIWYG Sep 12 '19 at 13:40
1
$\begingroup$
  1. bambara groundnut does not have a reference genome

This is your biggest problem if you want to do any sequencing based analysis. Neither GWAS nor RNAseq data analysis are possible without a reference genome (or at least transcriptome).
Depending on the availability, quality & similarity of reference genomes from related species it may however be possible to construct a reference transcriptome for your species from RNA sequencing data you get from RNAseq experiments. I'm whether it's feasible (or indeed possible) to try the same with genome sequencing data, since with only 10 individuals/strains you may only be able to get huge linkage groups for GWAS analysis.

Either way I'd highly recommend you to talk with someone (i.e. from your institute) who has experience in doing RNAseq or GWAS on plants, when planning this. (The experience with platns important, mammalian/model system genomes are much better studied and generally have less genome duplications, so analysis of plant data comes with its own set of baggage)

  1. Bearing in mind that the whole 95 varieties characterized were not subjected to water stress, do i need to set up another experiment entirely for the water stress and go ahead to make use of RNASeq to get the genes?

If you want to use RNAseq to find genes that are differentially expressed between states of water stress and no water stressin the same plant/species you will need to compare samples taken from corresponding conditions. This is only possible with RNAseq data.
Another approach could be to do some other experiments to determine which plants show a phenotype when exposed to water stress, and then look for correlations of this phenotype with (sequence) isoforms variants [GWAS like approach, possible with RNAseq or genomic sequencng data] or with expression levels of certain genes [again only possible with RNAseq data]. In thsi approach you need to be aware that your sequencing data either represents the ground state of all plants (no stress) or a potentially stressed state and missing one of these may mean that you can't find genes that change expression levels based on stress.

  1. It is obvious that some genes will be upregulated and some will be downregulated, how do i identify these genes and how do i know if they are expressed at the onset of stress, during, or after stress?

First, while most analysis tools will always give you some values for up- or downregulation, that does definitely not mean that these values are always meaningful. Only proper statistical testing can tell you which up/downregulation is somewhat reliable (it still wont tell you if its biologically meaningful) and depending on the quality of your data and the actual biological differences you may not be able to find anything.
Secondly, if you care about the timepoint of expression for specfic genes, you will need to some time series experiments. (Doing that for the whole RNAseq/GWAS would be very expensive, so ideally you first identify some candidate genes and then study their specific expression profiles with easier methods).

  1. when is the appropriate time to take my samples, i was considering taking at 50% flowering stage.

If you are not an expert on the species you#re working with, I'd again highly recomment you to speak with someone who is.

$\endgroup$
  • $\begingroup$ Many thanks, these points are highly insightful for me a start, i will discuss with someone that have experience to see what can be done. Thank you so much. $\endgroup$ – Samuel Sep 12 '19 at 12:04

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.