Methods For Comparing Transcriptomes?
1
2
Entering edit mode
13.2 years ago
Jared ▴ 20

My professor that I do research with has just proposed a new project for me, but if I accept we need to get this thing done in around 5 weeks so I want to know what it will entail before I sign on. He wants to identify the gene responsible for regulating flower color (purple vs white) in a species of mustard plant, but we know that the actual gene(s) coding for pigment production is identical in white vs purple plants (just not transcribed in white-flowered plants). He thinks the difference is in a mutation in some kind of regulatory gene because there aren't any other major differences in gene transcription. Our data set consists of four transcriptomes, each derived from about five individuals, from white and purple plants in two different populations. We're thinking we should be able to go through and find sites where white and purple-flowered plants differ but are the same among both populations, but he doesn't know exactly how to do this. I'm sure that I could write some kind of script to do this in an approximate manner but I feel like the real thing would use some kind of algorithm actually based off of proper math and statistics. Does anyone know of any methods to accomplish this sort of thing?

transcriptome • 4.8k views
ADD COMMENT
4
Entering edit mode

Do flower colors segregate in a Mendelian fashion, with complete penetrance? That would provide evidence that there is a single locus controlling color. Failing that, what is the evidence that this is a single mutation? I suspect you're signing up for more than 5 weeks worth of work.

ADD REPLY
0
Entering edit mode

Do you have a sequenced genome for the plant? What type of sequencing do you have?

ADD REPLY
0
Entering edit mode

Do you know the gene responsible for flower colour? And want to find its regulator? Microarray? rna-seq?

ADD REPLY
0
Entering edit mode

Hey! maybe RNAi is playing a role in the colour regulation. In fact RNAi was discovered in experiments concerning flower colors in petunias (http://en.wikipedia.org/wiki/RNA_interference#History_and_discovery), so I'm afraid it must be quite common in plants.

I agree with @David Quigley that this seems more than a 5-weeks work

ADD REPLY
3
Entering edit mode
13.2 years ago

Putting a time target of five weeks on completing a task that neither you nor your supervisor have ever done is unrealistic. What you can do in five weeks is:

  1. evaluate your data, probe it with various qualitative measures,
  2. quantify the expression levels that you can detect across the genes
  3. summarize the many genomic variations that your data will show

Once you do this - an in fact even these somewhat routine tasks above may take a lot longer than five weeks you will be in better position to determine if the original hypothesis is realistic in the first place.

I can't tell you just how many times I was asked to test all kinds of complex hypotheses only to find early on that there were flaws in data collection, experimental setup and/or sample preparation or sequencing that made the original goals unattainable. My advice is to first evaluate the data for a couple of weeks and try to answer a few simple questions first. If those all work out your can move to the second phase.

ADD COMMENT

Login before adding your answer.

Traffic: 1375 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6