Investigators at the University of Cambridge and the Wellcome Sanger Institute outline a sequence variation graph mapping strategy for more clearly spelling out small insertion and deletions (indels) in ancient DNA sequences. While ancient DNA molecules tend to be shorter than modern DNA, with specific chemical modifications that accumulate during the DNA degradation process, the team explains, the new results suggest that software that incorporates variation graphing to dial down potential biases introduced by reference sequences could boost the detection of indels and other variants in real and simulated ancient DNA datasets. "Our findings demonstrate that aligning [ancient DNA] sequences to variation graphs effectively mitigates the impact of reference bias when analyzing [ancient DNA]," they write, "while retaining mapping sensitivity and allowing detection of variation in particular indel variation that was previously missed."
A team from the University of California, Riverside, and the La Jolla Institute for Immunology report on a method called Mustache for using Hi-C and Micro-C map-based interaction data to track down chromatin loops. The approach "employs scale-space theory, a technical advance in computer vision, to detect blob-shaped objects in contact maps," the researchers note, adding that the chromatin loops reported by Mustache appear to line up across replicates and in datasets done with either Hi-C or Micro-C approaches. Moreover, they say, chromatin loops uncovered with the method lined up with those found using other calling approaches based on other types of interaction data. "Based on the results presented here, we believe that Mustache will become an essential tool in the analysis of high-resolution Hi-C and Micro-C contact," the authors write, "which are being produced in large numbers by the 4D Nucleosome project and other efforts."
Researchers from the University of Montpellier present software designed for quickly and accurately searching for disease-related features in large sequence sets. The "interactive multi-objective k-mer analysis," or iMOKA, method relies on a feature reduction step to reduce search space in large sequence sets, the team writes, while making it possible to assess data from more than one experiment together. The authors applied iMOKA to clinical sequence data from two breast cancer cohorts, a set of ovarian cancer sequences, and a collection of diffuse large B-cell lymphoma data, for example, and report that the software "found features that are more accurate than classical bioinformatics approaches." Moreover, they say, iMOKA "takes up less space, uses less memory, has faster runtimes, and can be run on a computer cluster or on a laptop."