A US Cancer Institute conducts research on individual genetic predisposition to cancer. Scientists in one of its laboratories suggested using DNA barcoding in their experiments. Statistical methods used for DNA barcode analysis require intensive computations, therefore the laboratory required custom bioinformatics software for this task.
The Axmor engineers developed software that analyzes DNA barcodes provided by researchers and compiles DNA barcode frequency matrices. Judging from these matrices, scientists can draw conclusions regarding a subject’s genetic predisposition to cancer.
To obtain more accurate results, the software compares DNA barcodes with ideal models generated before the experiment and corrects errors in the barcodes to be analyzed.
When elaborating the logic of DNA barcode analysis, we took advantage of various statistical methods used in computational biology for advanced data analysis and prediction. In particular, our developers extensively applied Markov chain Monte-Carlo methods, resampling, false discovery rate, and other techniques.
The software can simultaneously analyze DNA barcodes from a number of experiments. For a convenient display of the results, statistics located in one frequency matrix are automatically divided among respective parallel experiments.