Home
Author: John Andrew Reiser Farrell
Lab: Marth Lab
Contact: JAndrewRFarrell@gmail.com
RUFUS offers completely reference free variant detection with the ability to detect all of the variation that mapping based methods are currently capable of. RUFUS analyzes the k-mer count distribution between two samples to create an accurate model of read coverage across the genome. This model can then be leveraged to detect SNVs, Insertions and deletions, and copy number variants between a pair of Illumina sequenced samples. RUFUS does not require the massive computing power, or multiple sequencing libraries, which are required by whole genome assembly methods. This makes RUFUS particularly useful for researchers working on organisms for which there is no reference or a very poor reference. Additionally by removing the reference from mutation detection, reference bias is eliminated improving detection in 3 ways; Variations that occur in DNA that is either missing from the reference or is unmappable can now be discovered, improved the detection of INDELLs as well as improving the detection of rare variations due to the lack of bias, and a massive improvement in specificity as mapping error has been completely eliminated which contributes the majority of false positive calls in mapping based method.