Wichmann, S. | bioRxiv Science

bioRxiv2020

Computational design of genes encoding completely overlapping protein domains: Influence of genetic code and taxonomic rank

Overlapping genes (OLGs) with long protein-coding overlapping sequences are often excluded by genome annotation programs, with the exception of virus genomes. A recent study used a novel algorithm to construct OLGs from arbitrary protein domain pairs and concluded that virus genes are best suited for creating OLGs, a result which fitted with common assumptions. However, improving sequence evaluation using Hidden Markov Models shows that the previous result is an artifact originating from dataset-database biases. When parameters for OLG design and evaluation are optimized we find that 94.5% of the constructed OLG pairs score at least as highly as naturally occurring sequences, while 9.6% of the artificial OLGs cannot be distinguished from typical sequences in their protein family. Constructed OLG sequences are also indistinguishable from natural sequences in terms of amino acid identity and secondary structure, while the minimum nucleotide change required for overprinting an overlapping sequence can be as low as 1.8% of the sequence. Separate analysis of datasets containing only sequences from either archaea, bacteria, eukaryotes or viruses showed that, surprisingly, virus genes are much less suitable for designing OLGs than bacterial or eukaryotic genes. An important factor influencing OLG design is the structure of the standard genetic code. Success rates in different reading frames strongly correlate with their code-determined respective amino acid constraints. There is a tendency indicating that the structure of the standard genetic code could be optimized in its ability to create OLGs while conserving mutational robustness. The findings reported here add to the growing evidence that OLGs should no longer be excluded in prokaryotic genome annotations. Determining the factors facilitating the computational design of artificial overlapping genes may improve our understanding of the origin of these remarkable genetic constructs and may also open up exciting possibilities for synthetic biology.

bioinformatics↗

bioRxiv2020

Testing methods of homeland detection using synthetic data

Two families of quantitative methods have been used to infer geographical homelands of language families: Bayesian phylogeography and the diversity method. Bayesian methods model how populations may have moved using a phylogenetic tree as a backbone, while the diversity method assumes that the geographical area where linguistic diversity is highest likely corresponds to the homeland. No systematic tests of the performances of the different methods in a linguistic context have so far been published. Here we carry out performance testing by simulating language families, including branching structures and word lists, along with speaker populations moving in space. We test six different methods: two versions of BayesTraits; the relaxed random walk model of BEAST 2; our own RevBayes implementations of a fixed rates and a variable rates random walk model; and the diversity method. As a result of the tests we propose a hierarchy of performance of the different methods. Factors such as geographical idiosyncrasies, incomplete sampling, tree imbalance, and small family sizes all have a negative impact on performance, but mostly across the board, the performance hierarchy generally being impervious to such factors.

evolutionary biology↗