Simple, Effective and Stochastic Method for Validation of Transcriptomic Data

 

Background: Sequencing of the transcriptome has revolutionized quantitative and qualitative analyzes of prokaryotic and eukaryotic organisms. In the field of research on medically assisted reproduction, the selection of embryos with the best potential for implantation is the main challenge for biologists. Several studies suggest that the genes involved in oocyte cell cross talk could represent biomarkers for the selection of embryos with the greatest potential for implantation. Variability of different sources was observed during the transcriptomic experiment. Thus, one could have reasonable doubt about the validity of these transcriptomic data to provide a reliable and robust predictive model of pregnancy. Event is to be predicted at best by maximizing the likelihood criterion that will be discussed later. This vector Y which takes only two modalities: Positive or negative pregnancy is a random binary categorical and dependent process that follows a bi-nomial probability distribution. In the present study, the data are composed of 21 computative biomarker genes analyzed by quantitative polymerase chain reaction of 102 cumulus cell samples from patients undergoing in vitro fertilization. Results: The Stochastic Likehood and the Youden’s Index results will make it possible to discard the data in the case where they are biased and thus save time in terms of processing. Conclusion: Indeed, the Stochastic Likehood and the Youden’s Index showed that the transcriptomic data used are biased