A probabilistic model for indel evolution: differentiating insertions from deletions

Persistent homology reveals strong phylogenetic signal in 3D protein structures

Changes that occur in proteins over time provide a phylogenetic signal that can be used to decipher their evolutionary history and the relationships between organisms. Sequence comparison is the most common way to access this phylogenetic signal, while those based on 3D structure comparisons are still in their infancy. In this study, we propose an effective approach based on Persistent Homology Theory (PH) to extract the phylogenetic information contained in protein structures. PH provides efficient and robust algorithms for extracting and comparing geometric features from noisy datasets at different spatial resolutions. PH has a growing number of applications in the life sciences, including the study of proteins (e.g. classification, folding). However, it has never been used to study the phylogenetic signal they may contain. Here, using 518 protein families, representing 22,940 protein sequences and structures, from 10 major taxonomic groups, we show that distances calculated with PH from protein structures correlate strongly with phylogenetic distances calculated from protein sequences, at both small and large evolutionary scales. We test several methods for calculating PH distances and propose some refinements to improve their relevance for addressing evolutionary questions. This work opens up new perspectives in evolutionary biology by proposing an efficient way to access the phylogenetic signal contained in protein structures, as well as future developments of topological analysis in the life sciences.

Effect of tokenization on transformers for biological sequences

MOTIVATION

Deep-learning models are transforming biological research, including many bioinformatics and comparative genomics algorithms, such as sequence alignments, phylogenetic tree inference, and automatic classification of protein functions. Among these deep-learning algorithms, models for processing natural languages, developed in the natural language processing (NLP) community, were recently applied to biological sequences. However, biological sequences are different from natural languages, such as English, and French, in which segmentation of the text to separate words is relatively straightforward. Moreover, biological sequences are characterized by extremely long sentences, which hamper their processing by current machine-learning models, notably the transformer architecture. In NLP, one of the first processing steps is to transform the raw text to a list of tokens. Deep-learning applications to biological sequence data mostly segment proteins and DNA to single characters. In this work, we study the effect of alternative tokenization algorithms on eight different tasks in biology, from predicting the function of proteins and their stability, through nucleotide sequence alignment, to classifying proteins to specific families.

RESULTS

We demonstrate that applying alternative tokenization algorithms can increase accuracy and at the same time, substantially reduce the input length compared to the trivial tokenizer in which each character is a token. Furthermore, applying these tokenization algorithms allows interpreting trained models, taking into account dependencies among positions. Finally, we trained these tokenizers on a large dataset of protein sequences containing more than 400 billion amino acids, which resulted in over a 3-fold decrease in the number of tokens. We then tested these tokenizers trained on large-scale data on the above specific tasks and showed that for some tasks it is highly beneficial to train database-specific tokenizers. Our study suggests that tokenizers are likely to be a critical component in future deep-network analysis of biological sequence data.

AVAILABILITY AND IMPLEMENTATION

Code, data, and trained tokenizers are available on https://github.com/technion-cs-nlp/BiologicalTokenizers.

Statistical framework to determine indel-length distribution

MOTIVATION

Insertions and deletions (indels) of short DNA segments, along with substitutions, are the most frequent molecular evolutionary events. Indels were shown to affect numerous macro-evolutionary processes. Because indels may span multiple positions, their impact is a product of both their rate and their length distribution. An accurate inference of indel-length distribution is important for multiple evolutionary and bioinformatics applications, most notably for alignment software. Previous studies counted the number of continuous gap characters in alignments to determine the best-fitting length distribution. However, gap-counting methods are not statistically rigorous, as gap blocks are not synonymous with indels. Furthermore, such methods rely on alignments that regularly contain errors and are biased due to the assumption of alignment methods that indels lengths follow a geometric distribution.

RESULTS

We aimed to determine which indel-length distribution best characterizes alignments using statistical rigorous methodologies. To this end, we reduced the alignment bias using a machine-learning algorithm and applied an Approximate Bayesian Computation methodology for model selection. Moreover, we developed a novel method to test if current indel models provide an adequate representation of the evolutionary process. We found that the best-fitting model varies among alignments, with a Zipf length distribution fitting the vast majority of them.

AVAILABILITY AND IMPLEMENTATION

The data underlying this article are available in Github, at https://github.com/elyawy/SpartaSim and https://github.com/elyawy/SpartaPipeline.

Simulations of Sequence Evolution: How (Un)realistic They Are and Why

MOTIVATION

Simulating multiple sequence alignments (MSAs) using probabilistic models of sequence evolution plays an important role in the evaluation of phylogenetic inference tools and is crucial to the development of novel learning-based approaches for phylogenetic reconstruction, for instance, neural networks. These models and the resulting simulated data need to be as realistic as possible to be indicative of the performance of the developed tools on empirical data and to ensure that neural networks trained on simulations perform well on empirical data. Over the years, numerous models of evolution have been published with the goal to represent as faithfully as possible the sequence evolution process and thus simulate empirical-like data. In this study, we simulated DNA and protein MSAs under increasingly complex models of evolution with and without insertion/deletion (indel) events using a state-of-the-art sequence simulator. We assessed their realism by quantifying how accurately supervised learning methods are able to predict whether a given MSA is simulated or empirical.

RESULTS

Our results show that we can distinguish between empirical and simulated MSAs with high accuracy using two distinct and independently developed classification approaches across all tested models of sequence evolution. Our findings suggest that the current state-of-the-art models fail to accurately replicate several aspects of empirical MSAs, including site-wise rates as well as amino acid and nucleotide composition.

Development of Omni InDel and supporting database for maize

Insertions-deletions (InDels) are the second most abundant molecular marker in the genome and have been widely used in molecular biology research along with simple sequence repeats (SSR) and single-nucleotide polymorphisms (SNP). However, InDel variant mining and marker development usually focuses on a single type of dimorphic InDel, which does not reflect the overall InDel diversity across the genome. Here, we developed Omni InDels for maize, soybean, and rice based on sequencing data and genome assembly that included InDel variants with base lengths from 1 bp to several Mb, and we conducted a detailed classification of Omni InDels. Moreover, we screened a set of InDels that are easily detected and typed (Perfect InDels) from the Omni InDels, verified the site authenticity using 3,587 germplasm resources from 11 groups, and analyzed the germplasm resources. Furthermore, we developed a Multi-InDel set based on the Omni InDels; each Multi-InDel contains multiple InDels, which greatly increases site polymorphism, they can be detected in multiple platforms such as fluorescent capillary electrophoresis and sequencing. Finally, we developed an online database website to make Omni InDels easy to use and share and developed a visual browsing function called "Variant viewer" for all Omni InDel sites to better display the variant distribution.

The Telomeric Repeats of HHV-6A Do Not Determine the Chromosome into Which the Virus Is Integrated

Human herpes virus 6A (HHV-6A) is able to integrate into the telomeric and subtelomeric regions of human chromosomes representing chromosomally integrated HHV-6A (ciHHV-6A). The integration starts from the right direct repeat (DRR) region. It has been shown experimentally that perfect telomeric repeats (pTMR) in the DRR region are required for the integration, while the absence of the imperfect telomeric repeats (impTMR) only slightly reduces the frequency of HHV-6 integration cases. The aim of this study was to determine whether telomeric repeats within DRR may define the chromosome into which the HHV-6A integrates. We analysed 66 HHV-6A genomes obtained from public databases. Insertion and deletion patterns of DRR regions were examined. We also compared TMR within the herpes virus DRR and human chromosome sequences retrieved from the Telomere-to-Telomere consortium. Our results show that telomeric repeats in DRR in circulating and ciHHV-6A have an affinity for all human chromosomes studied and thus do not define a chromosome for integration.

Analysis of Copy Number Variation in the Whole Genome of Normal-Haired and Long-Haired Tianzhu White Yaks

Long-haired individuals in the Tianzhu white yak population are a unique genetic resource, and have important landscape value. Copy number variation (CNV) is an important source of phenotypic variation in mammals. In this study, we used resequencing technology to detect the whole genome of 10 long-haired Tianzhu white yaks (LTWY) and 10 normal-haired Tianzhu white yaks (NTWY), and analyzed the differences of CNV in the genome of LTWYs and NTWYs. A total of 110268 CNVs were identified, 2006 CNVRs were defined, and the distribution map of these CNVRs on chromosomes was constructed. The comparison of LTWYs and NTWYs identified 80 differential CNVR-harbored genes, which were enriched in lipid metabolism, cell migration and other functions. Notably, some differential genes were identified as associated with hair growth and hair-follicle development (e.g., ASTN2, ATM, COL22A1, GK5, SLIT3, PM20D1, and SGCZ). In general, we present the first genome-wide analysis of CNV in LTWYs and NTWYs. Our results can provide new insights into the phenotypic variation of different hair lengths in Tianzhu white yaks.

The evolutionary dynamics that retain long neutral genomic sequences in face of indel deletion bias: a model and its application to human introns

Insertions and deletions (indels) of short DNA segments are common evolutionary events. Numerous studies showed that deletions occur more often than insertions in both prokaryotes and eukaryotes. It raises the question why neutral sequences are not eradicated from the genome. We suggest that this is due to a phenomenon we term border-induced selection. Accordingly, a neutral sequence is bordered between conserved regions. Deletions occurring near the borders occasionally protrude to the conserved region and are thereby subject to strong purifying selection. Thus, for short neutral sequences, an insertion bias is expected. Here, we develop a set of increasingly complex models of indel dynamics that incorporate border-induced selection. Furthermore, we show that short conserved sequences within the neutrally evolving sequence help explain: (i) the presence of very long sequences; (ii) the high variance of sequence lengths; and (iii) the possible emergence of multimodality in sequence length distributions. Finally, we fitted our models to the human intron length distribution, as introns are thought to be mostly neutral and bordered by conserved exons. We show that when accounting for the occurrence of short conserved sequences within introns, we reproduce the main features, including the presence of long introns and the multimodality of intron distribution.

An Approximate Bayesian Computation Approach for Modeling Genome Rearrangements

The inference of genome rearrangement events has been extensively studied, as they play a major role in molecular evolution. However, probabilistic evolutionary models that explicitly imitate the evolutionary dynamics of such events, as well as methods to infer model parameters, are yet to be fully utilized. Here, we developed a probabilistic approach to infer genome rearrangement rate parameters using an Approximate Bayesian Computation (ABC) framework. We developed two genome rearrangement models, a basic model, which accounts for genomic changes in gene order, and a more sophisticated one which also accounts for changes in chromosome number. We characterized the ABC inference accuracy using simulations and applied our methodology to both prokaryotic and eukaryotic empirical datasets. Knowledge of genome-rearrangement rates can help elucidate their role in evolution as well as help simulate genomes with evolutionary dynamics that reflect empirical genomes.

Ancestral Sequence Reconstruction for Exploring Alkaloid Evolution

The complex and bioactive monoterpene indole alkaloids (MIAs) found in Catharanthus roseus and related species are the products of many millions of years of evolution through mutation and natural selection. Ancestral sequence reconstruction (ASR) is a method that combines phylogenetic analysis and experimental biochemistry to infer details about past events in protein evolution. Here, I propose that ASR could be leveraged to understand how enzymes catalyzing the formation of complex alkaloids arose over evolutionary time. I discuss the steps of ASR, including sequence selection, multiple sequence alignment, tree inference, and the generation and characterization of inferred ancestral enzymes.