I tested 189 faba bean lines in field with two replications in one year and assessed a number of traits. These lines were genotyped using SNPs (156).
Do you think that I can do association analysis for these traits? and
Do you think one year and two replications are applicable to do such analysis?
The number of genotypes (189) is good for association analysis. The 156 SNPs is also not a bad number, but more the markers better will be your confidence or reliability of association analysis and possibility to identify the potential gene for the trait of your interest. But i am concerned about the two replications. You may need to have at-least three replications. Statistically speaking three reps are necessary for most of the analysis and to publish your research in peer-reviewed journals. One year data is enough for doing association analysis but atleast two year data is needed for publishing your research in peer-reviewed journals
I think you should move a head and do an association analysis on the above data and see what you end up. Sometimes there may be some good chance to identify the chromosomal location for you marker-trait association which would still be useful for further analysis.
Thank you so much for your answer, I have the same impression but I wanted to check from others.
As Arun Prabhu Dhanapal said, replications are really concern for doing analysis. Even if you have three reps of single year, you will also feel uncomfortable with your confused data. So, as per my SNP data and phenotyping data I do have some confusion, too. Others researchers has given sugestions to do two year experiement. So, my advise you could go with your data and see what you get with your analysis, but best is use of three reps and two years data.
I already did the association analysis for such data and I found some putative qtls for day of flowering , 1000 kernel weight and others
Ahmed,
one year data are not sufficient to prove that you QTLs are really linked to the traits, next year data can give completly different result.
2 reps are bad from all points of veiw - 4 is better, 3 - is so-so.
dear Ahmed as the previous scientists said, data with two replications is not quite enough to get a reliable association mapping, especially when you are dealing with quantitative trait loci since they are highly influenced by environment and of low heritability. therefore repeating the testing and phenotyping for at least two years or even more you will get reliable and more accurate results
Hi., I feel its better to go for validation of those marker-trait analysis and it advantageous for you to publish paper in good journal since you have only two replications and data for one season.
Thanks for the valuable discussion. The problem I do not have a chance to do it for a second year. I would like to publish the results even if in journal with less IF. The traits "Awadalla" showed high heritability more that 0.75. For all of you, do you think that I can name this part " preliminary association analysis .............. " . Just to be more precious. I look forward to hearing from you.
The problem is that one year, one location, two reps data must have high probability of environmental effect on association strength and high experiment error. For any journal, editor will not accept such publication. Can you repeat the test in glasshouse?
Alex, I fully agree with you and all prestigious researchers above. I just wanted to share experiences and I may find someone had a similar case.
Ahmed, you can report such data as preliminary in conference poster or abstact (some accept papers up to 4 pages) or write a draft of paper and place it at http://arxiv.org/archive/astro-ph, but maybe, it would be better to find a student/researcher who will continue your experiment for one more season?
- Badges
- Science topic
- Similar topics
- Correction