A gene-based model of fitness and its implications for genetic variation

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Сен. 14, 2024

A widely used model of the effects mutations on fitness (the "sites" model) assumes that heterozygous recessive or partially deleterious at different sites in a gene complement each other, similarly to genes. However, general lack complementation between major effect allelic suggests an alternative possibility, which we term "gene" model. This pair trans behave effectively as homozygotes, so fitnesses heterozygotes are lower than those cis heterozygotes. We examine properties two models, using both analytical and simulation methods. show predicts positive linkage disequilibrium (LD) variants within coding sequence, under conditions when zero slightly negative LD. also focussing rare examining patterns LD, especially with Lewontin's D ' measure, is likely produce misleading results respect inferences concerning causes sign Synergistic epistasis pairs was modeled; it less LD The theoretical discussed relation natural populations several species.

Язык: Английский

---

Performance evaluation of adaptive introgression classification methods

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Июнь 14, 2024

Abstract Introgression, the incorporation of foreign variants through hybridization and repeated backcross, is increasingly being studied for its potential evolutionary consequences, one which adaptive introgression (AI). In recent years, several statistical methods have been proposed detection loci that undergone introgression. Most these tested developed to infer presence Neanderthal or Denisovan AI in humans. Currently, behaviour when faced with genomic datasets from scenarios other than human lineage remains unknown. This study therefore focuses on testing performance using test data sets simulated under various inspired by history human, wall lizard ( Podarcis ) bear Ursus lineages. These lineages were chosen represent different combinations divergence migration times. We impact parameters, as well rate, population size, selection coefficient recombination hotspots, three (VolcanoFinder, Genomatnn MaLAdapt) a standalone summary statistic (Q95( w , y )). Furthermore, hitchhiking effect an adaptively introgressed mutation can strong flanking regions, discrimination between windows classes i.e. AI/non-AI). For this reason, types non-AI are taken into account our analyses: independently neutral windows, adjacent window AI, coming second chromosome unlinked AI. Our results highlight importance taking training order correctly identify Finally, tests show based Q95 seem be most efficient exploratory

Язык: Английский

---

Digital image processing to detect adaptive evolution

Molecular Biology and Evolution, Год журнала: 2024, Номер 41(12)

Опубликована: Ноя. 20, 2024

In recent years, advances in image processing and machine learning have fueled a paradigm shift detecting genomic regions under natural selection. Early techniques employed population-genetic summary statistics as features, which focus on specific patterns expected by adaptive neutral processes. Though such engineered features are important when training data limited, the ease at simulated can now be generated has led to development of approaches that take representations haplotype alignments automatically extract using convolutional neural networks. Digital methods termed α-molecules class for multiscale representation objects diverse set from images. One α-molecule method, wavelet decomposition, lends greater control over high-frequency components Another curvelet is an extension concept considers events occurring along curves within We show application these yield high true positive rate accuracy detect hard soft selective sweep signatures with both linear nonlinear classifiers. Moreover, we find models easy visualize interpret, performance rivaling those contemporary deep sweeps.

Язык: Английский

---

Evolutionary rescue model informs strategies for driving cancer cell populations to extinction

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Ноя. 28, 2024

Cancers exhibit a remarkable ability to develop resistance range of treatments, often resulting in relapse following first-line therapies and significantly worse outcomes for subsequent treatments. While our understanding the mechanisms dynamics emergence during cancer therapy continues advance, questions remain about how minimize probability that will evolve, thereby improving long-term patient outcomes. Here, we present an evolutionary simulation model clonal population cells can acquire mutations one or more We leverage this examine efficacy two-strike "extinction therapy" protocol, which two treatments are applied sequentially first contract vulnerable state then push it extinction, compare combination protocol. investigate factors such as timing switch between strikes, rate resistant mutations, doses drugs, presence cross-resistance, whether is binary quantitative trait affect outcome. Our results show switching second strike has marked effect on likelihood driving extinction outperforms when cross-resistance present. conduct silico trial reveals why succeed fail. Finally, demonstrate conclusions hold quantitative, multi-locus trait.

Язык: Английский

---