SeqTar users: is it right to calculate the probability of the valid peak using the binomial probability density function as they tell in the paper (http://nar.oxfordjournals.org/content/early/2011/12/02/nar.gkr1092.full)?
Considering that the binomial PDF for a certain value returns the probability to get a peak *exactly the same height* of the one under analysis, it seems to me that the right thing to do would be instead to calculate 1- (the binomial cumulative distribution for height of the peak under analysis-1).
This would give the probability of getting a peak *the same height of the valid peak OR MORE* given the trials and the probability, as before. Indeed using the probability density function might create problems: if a gene is heavily covered by PARE reads and the valid peak has much less reads than expected, the probability might be very low while the peak is not outstanding with respect to the surrounding nucleotides.
Using the cumulative probability, the probability of getting higher peaks would approach one, as expected. Is it right?
hi,
The P-value of valid reads in SeqTar is calculated by 1 - binocdf(k,n,q), where k = = max(n9, n10, n11), n9 to n11 were the number of reads mapped to the positions
opposite to the 9th to 11th nt of the sRNA, respectively, n was the total number of reads
that were mapped to the whole target sequence, and q was a constant that stands for the probability that a mapped read was from any nucleotide of the target sequence.
I noticed the Equation 1 in the manuscript lacks a sum_{i=k}^{n}.
The implementation of the SeqTar program use 1 - binocdf(k,n,q). Thus, the results are reliable and correct.
Hope these clarify your concern.
All the best.
sincerely
zheng, yun
Associate Professor, PI
Faculty of Life Science and Technology
Kunming University of Science and Technology
727 South Jingming Road
Kunming, Yunnan, China 650500
Tel:86-871-65920570
Fax:86-871-65920570
Email:[email protected];[email protected]
web: http://www.researcherid.com/rid/I-7011-2012;http://www.researchgate.net/profile/Yun_Zheng
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biomolecules
- Nucleic Acids
- RNA
- Small RNA
- microRNA