Effective p-value computations using Finite Markov Chain Imbedding
Author Information
Author(s): Nuel Grégory
Primary Institution: Laboratoire Statistique et Génome, UEVE, CNRS (8071), INRA (1152), Evry, France
Hypothesis
Can we compute exact p-values for local scores and pattern statistics using Finite Markov Chain Imbedding (FMCI)?
Conclusion
The new FMCI algorithms provide exact p-values for local scores and pattern statistics, outperforming previous approximation methods.
Supporting Evidence
- The new algorithms allow for the first time the computation of exact p-values for practical studies.
- Asymptotic approximations were found to be unreliable for 99.5% of the sequences considered.
- The FMCI algorithms are more reliable, easier to implement, and faster than previous methods.
Takeaway
This study shows a new way to calculate p-values accurately for patterns in sequences, which helps scientists understand proteins better.
Methodology
The study developed recursive algorithms to compute exact cumulative distribution functions and p-values for local scores and pattern statistics.
Limitations
The algorithms may require significant computational time for very large datasets.
Participant Demographics
The study analyzed approximately 200,000 protein sequences from the Swissprot database.
Statistical Information
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website