An article from Loschmidt Laboratories has been ranked among the 10% most cited articles in PLOS Computational Biology in 2016
The article, which describes the PredictSNP portal, was created by a team from FNUSA-ICRC composed of experts from the Loschmidt laboratories at MU and BUT and has proved a real publication success. Congratulations!
This successful article, by Bendl, J., Musil, M., Stourac, J., Zendulka, J., Damborsky, J. and Brezovsky, J, is entitled PredictSNP2: A Unified Platform for Accurately Evaluating SNP Effects by Exploiting the Different Characteristics of Variants in Distinct Genomic Regions.
The PredictSNP portal can help identify whether a deviation in the human genome will result in any of six to eight thousand rare diseases caused by mutations at a single site in human DNA. These diseases include, cystic fibrosis, Huntington’s disease and spinal muscular atrophy, which affect about one percent of the population.
The abbreviation SaNP in the portal name stands for “single nucleotide polymorphism”, which refers to the range of Mendelian diseases arising from a mutation in a single part of the individual’s DNA. “the first version of the portal that we created in 2014 dealt with mutations in genes that make up only about 1.5% of genomic DNA. We have now extended the portal, making it possible to examine the consequences of changes in the whole genome", explained Jaroslav Bendl, one of the authors of the new portal who worked at MU’s Loschmidt Laboratories during his studies and now works at Mount Sinai School of Medicine in New York.
The first author of the article, J. Bendl, now works at the prestigious Mount Sinai School of Medicine in New York.
Photo by Ema Wiesnerová. Magazine M: Masaryk University monthly. Available here under CC BY 3.0 CZ, license conditions are available at http://creativecommons.org/licenses/by/3.0/cz/ legal code.
While there are similar portals around the world, independent comparative studies show that the performance parameters they declare are often overestimated due to the use of incorrect testing techniques. Many of them also lack information on the estimated reliability of the prediction, such that the user has no guidance for estimating the ‘degree of harmfulness’ of the mutation investigated. “PredictSNP solves these problems by assigning its own weightings derived from our own testing of these tools compared with the predictions provided for each tool. Using these weightings, which are specific to different regions of the genome, PredictSNP then combines the prediction tools together, leading to an improvement in predictive success of up to five percent”, Jaroslav explained.
Experts continue to work on the portal as it is especially important to regularly update the set of integrated prediction programs and databases of known diseases. “In the future, we would also like to focus on predictions for some neurodegenerative diseases, such as Alzheimer’s or Parkinson's disease”, added Jiří Damborský, head of the research team at Loschmidt Laboratories.
The portal is freely available at: http://loschmidt.chemi.muni.cz/predictsnp/.
