Biology subjects

Pagnani, A.

Publications and source records attributed to Pagnani, A..

16 recordsLinked to original sources

bioRxiv2026

Modeling Protein Sequence Evolution as an Ornstein-Uhlenbeck Process in a Latent Space

High-throughput directed evolution produces longitudinal sequence libraries that are ideal for probing local fitness neighborhoods but often underpowered for global inference tasks such as contact prediction. We present an unsupervised inference model that integrates directed-evolution sequencing time series with natural homologs. We project sequences into a low-dimensional latent space learned from the natural multiple sequence alignment and model the experimental process as an Ornstein-Uhlenbeck dynamics in that space. Maximum-likelihood estimation of the latent drift and noise parameters determines a stationary Gaussian distribution, which induces an effective Potts model in sequence space. The inferred couplings improve structural contact prediction by combining global evolutionary constraints from nature with local, experiment-specific signals. Experiments on PSE1 {beta}-lactamase and dihydrofolate reductase demonstrate the ability to identify correct complementary contacts not recovered by methods using either natural or experimental data alone, with gains concentrated in intermediate- and long-range contacts.

bioinformatics↗

bioRxiv2026

Partner determination from protein sequences using class information with CLAPP

Protein-protein interactions underpin nearly all cellular processes, making their accurate identification a central challenge in biology. With the rapid expansion of genomic data, sequence-based computational approaches have emerged as a powerful route to infer such interactions, complementing experimental methods that are often prohibitively time- and resource-intensive. This challenge becomes particularly acute in the presence of paralogs, which arise through gene duplication and typically diversify toward distinct, though sometimes overlapping, functions. Reconstructing their interaction networks is therefore essential for understanding a wide range of biological processes. Protein paralogs within a family can often be subdivided into classes based on a range of properties, including functional, structural and architectural features. When interactions between these classes are conserved across organisms, such that sequences from one class interact exclusively with sequences from another, this information can be used to solve the paralog matching problem. We introduce here CLAPP (CLAss Pooling for Paralog matching), a method for predicting interacting paralogs by pooling interaction scores from different subclasses across organisms. We apply it to scores extracted using coevolution-based methods. Pooling scores at the class level reduces noise in the interaction scores and replaces organism-specific assignments with a single shared assignment, improving performance and substantially reducing computational cost. We apply CLAPP to bacterial systems including histidine kinases and response regulators, as well as interacting families of chaperones and co-chaperones, and recover known interaction partners.

bioinformatics↗

bioRxiv2026

Modeling Protein Evolution via Generative Inference From Monte Carlo Chains to Population Genetics

Generative models derived from large protein sequence alignments define complex fitness landscapes, but their utility for accurately modeling non-equilibrium evolutionary dynamics remains unclear. In this work, we perform a rigorous comparative analysis of three simulation schemes, designed to mimic evolution in silico by local sampling of the probability distribution defined by a generative model. We compare standard independent Markov Chain Monte Carlo, Monte Carlo on a phylogenetic tree, and a population genetics dynamics, benchmarking their outputs against deep sequencing data from four distinct in vitro evolution experiments. We find that standard Monte Carlo fails to reproduce the correct phylogenetic structure and generates unrealistic, gradual mutational sweeps. Performing Monte Carlo on a tree inferred from data improves phylogenetic fidelity and historical accuracy. The population genetics scheme successfully captures phylogenetic correlations, mutational abundances, and selective sweeps as emergent properties, without the need to infer additional information from data. However, the latter choice come at the price of not sampling the proper generative model distribution at long times. Our findings highlight the crucial role of phylogenetic correlations and finite-population effects in shaping evolutionary trajectories on fitness landscapes. These models therefore provide powerful tools for predicting complex adaptive paths and for reliably extrapolating evolutionary dynamics beyond current experimental limitations.

evolutionary biology↗

bioRxiv2025

Generative continuous time model reveals epistatic signatures in protein evolution

Protein evolution is fundamentally shaped by epistasis, where the effect of a mutation depends on the sequence context. As standard phylogenetic methods assume independently evolving sites, there is a need for more complex models based on accurate estimations of the fitness landscape. Good candidates are modern generative models - such as the Potts model - which successfully capture epistatic effects. However, recent works on generative evolutionary models usually use discrete time, making them difficult to integrate with the standard frameworks in evolutionary biology. We introduce a continuous-time sequence evolution model using the Gillespie algorithm and parameterized by a generative Potts model. This approach enables us to simulate realistic, family-specific evolutionary trajectories and allows for direct comparison with independent-site models. Surprisingly, we find that while epistasis significantly slows down evolution, it does not change the average evolutionary rates at individual sites. This is explained by the rate heterogeneity caused by context-dependence: we show that the rate at some positions varies between null to high values depending on the context, while other positions are essentially independent from the context. Finally, we show that epistasis leads to a systematic underestimation bias in the inference of evolutionary distance between sequences. Overall, our work provides a new tool for simulating realistic protein evolution and offers novel insights into the complex interplay between epistasis and evolutionary dynamics. Significance statementUnderstanding how proteins evolve is central to molecular biology and phylogenetics. Traditional evolutionary models assume that mutations act independently at each position in a sequence. This neglects epistasis -- the fact that the effect of a mutation depends on the rest of the sequence -- which is known to be ubiquitous in proteins. By simulating protein evolution in continuous time using a generative model, our approach produces realistic sequences and reveals how epistasis shapes evolutionary dynamics. We find that epistasis slows down evolution and can mislead common methods for estimating evolutionary timescales. This work bridges modern generative models of proteins and phylogenetics, providing new tools to better understand molecular evolution.

bioinformatics↗

bioRxiv2025

Controllable Protein Design via Autoregressive Direct Coupling Analysis Conditioned on Principal Components

We present FeatureDCA, a statistical framework for protein sequence modeling and generation that extends Direct Coupling Analysis (DCA) with biologically meaningful conditioning. The method can leverage different kinds of information, such as phylogeny, optimal growth temperature, enzymatic activity or, as in the case presented here, principal components derived from multiple sequence alignments, and use it to improve the learning process and consequently efficiently condition the generative process. FeatureDCA allows sampling to be guided toward specific regions of sequence space while maintaining the efficiency and interpretability of Potts-based inference. Across multiple protein families, our autoregressive implementation of FeatureDCA matches or surpasses the generative accuracy of established models in reproducing higher-order sequence statistics while preserving substantial sequence diversity. Structural validation with AlphaFold and ESMFold confirms that generated sequences adopt folds consistent with their intended wild-type targets. In a detailed case study of the Response Regulator family (PF00072), which comprises distinct structural subclasses linked to different DNA-binding domains, FeatureDCA accurately reproduces class-specific architectures when conditioned on subtype-specific principal components, highlighting its potential for fine-grained structural control. Predictions of experimental deep mutational scanning data show accuracy comparable to that of unconditioned autoregressive Potts models, indicating that FeatureDCA also captures local functional constraints. These results position FeatureDCA as a flexible and transparent approach for targeted sequence generation, bridging statistical fidelity, structural realism, and interpretability in protein design.

bioinformatics↗

bioRxiv2024

Reconstruction of ancestral protein sequences using autoregressive generative models

Ancestral sequence reconstruction (ASR) is an important tool to understand how protein structure and function changed over the course of evolution. It essentially relies on models of sequence evolution that can quantitatively describe changes in a sequence over time. Such models usually consider that sequence positions evolve independently from each other and neglect epistasis: the context-dependence of the effect of mutations. On the other hands, the last years have seen major developments in the field of generative protein models, which learn constraints associated with structure and function from large ensembles of evolutionarily related proteins. Here, we show that it is possible to extend a specific type of generative model to describe the evolution of sequences in time while taking epistasis into account. We apply the developed technique to the problem of Ancestral Sequence Reconstruction (ASR): given a protein family and its evolutionary tree, we try to infer the sequences of extinct ancestors. Using both simulations and data coming from experimental evolution we show that our method outperforms state-of-the-art ones. Moreover, it allows for sampling a greater diversity of potential ancestors, allowing for a less biased characterization of ancestral sequences.

bioinformatics↗

bioRxiv2024

Direct Coupling Analysis and The Attention Mechanism

Proteins are involved in nearly all cellular functions, encompassing roles in transport, signaling, enzymatic activity, and more. Their functionalities crucially depend on their complex three-dimensional arrangement. For this reason, being able to predict their structure from the amino acid sequence has been and still is a phenomenal computational challenge that the introduction of AlphaFold solved with unprecedented accuracy. However, the inherent complexity of AlphaFolds architectures makes it challenging to understand the rules that ultimately shape the proteins predicted structure. This study investigates a single-layer unsupervised model based on the attention mechanism. More precisely, we explore a Direct Coupling Analysis (DCA) method that mimics the attention mechanism of several popular Transformer architectures, such as AlphaFold itself. The models parameters, notably fewer than those in standard DCA-based algorithms, can be directly used for extracting structural determinants such as the contact map of the protein family under study. Additionally, the functional form of the energy function of the model enables us to deploy a multi-family learning strategy, allowing us to effectively integrate information across multiple protein families, whereas standard DCA algorithms are typically limited to single protein families. Finally, we implemented a generative version of the model using an autoregressive architecture, capable of efficiently generating new proteins in silico. The effectiveness of our Attention-Based DCA architecture is evaluated using different families of evolutionary-related proteins, whose structural data is sourced from the Pfam database.

bioinformatics↗

bioRxiv2023

Unsupervised modeling of mutational landscapes of adeno-associated viruses viability

Adeno-associated viruses 2 (AAV2) are minute viruses renowned for their capacity to infect human cells and akin organisms. They have recently emerged as prominent candidates in the field of gene therapy, primarily attributed to their inherent non-pathogenic nature in humans and the safety associated with their manipulation. The efficacy of AAV2 as gene therapy vectors hinges on their ability to infiltrate host cells and subsequently replicate within them, a phenomenon reliant on their competence to construct a capsid capable of breaching the nucleus of the target cell. To enhance their infection potential, researchers have extensively scrutinized various combinatorial libraries by introducing mutations into the capsid, aiming to boost their effectiveness. The emergence of high-throughput experimental techniques, like Deep Mutational Scanning (DMS), has made it feasible to experimentally assess the fitness of these libraries for their intended purpose. Notably, machine learning is starting to demonstrate its potential in addressing predictions within the mutational landscape from sequence data. In this context, we introduce a biophysically-inspired model designed to predict the viability of genetic variants in DMS experiments. This model is tailored to a specific segment of the CAP region within AAV2s capsid protein. To evaluate its effectiveness, we conduct model training with diverse datasets, each tailored to explore different aspects of the mutational landscape influenced by the selection process. Our assessment of the biophysical model centers on two primary objectives: (i) providing quantitative forecasts for the log-selectivity of variants and (ii) deploying it as a binary classifier to categorize sequences into viable and non-viable classes.

biophysics↗

bioRxiv2023

Inference and design of antibody specificity: from experiments to models and back

Exquisite binding specificity is essential for many protein functions but is difficult to engineer. Many biotechnological or biomedical applications require the discrimination of very similar ligands, which poses the challenge of designing protein sequences with highly specific binding profiles. Current methods for generating specific binders rely on in vitro selection experiments, but these have limitations in terms of library size and control over specificity profiles. We present a multi-stage approach that overcomes these limitations by combining high-throughput sequencing of phage display experiments with machine learning and biophysical modeling. Our models predict the binding profiles of antibodies against multiple ligands and generate antibody sequences with desired specificity profiles. The approach involves the identification of different binding modes, each associated with a particular ligand against which the antibodies are either selected or not. We demonstrate that the model successfully disentangles these modes, even when they are associated with chemically very similar ligands. Additionally, we demonstrate and validate experimentally the computational design of antibodies with customized specificity profiles, either with specific high affinity for a particular target ligand, or with cross-specificity for multiple target ligands. Overall, our results showcase the potential of leveraging a biophysical model learned from selections against multiple ligands to design proteins with tailored specificity, with applications to protein engineering extending beyond the design of antibodies.

biophysics↗

bioRxiv2023

Inference of annealed protein fitness landscapes with AnnealDCA

The design of proteins with specific tasks is a major challenge in molecular biology with important diagnostic and therapeutic applications. High-throughput screening methods have been developed to systematically evaluate protein activity, but only a small fraction of possible protein variants can be tested using these techniques. Computational models that explore the sequence space in-silico to identify the fittest molecules for a given function are needed to overcome this limitation. In this article, we propose AnnealDCA, a machine-learning framework to learn the protein fitness landscape from sequencing data derived from a broad range of experiments that use selection and sequencing to quantify protein activity. We demonstrate the effectiveness of our method by applying it to antibody Rep-Seq data of immunized mice and screening experiments, assessing the quality of the fitness landscape reconstructions. Our method can be applied to most experimental cases where a population of protein variants undergoes various rounds of selection and sequencing, without relying on the computation of variant enrichment ratios, and thus can be used even in cases of disjoint sequence samples.

biophysics↗

bioRxiv2022

Context-Aware Generative Models for Multi-Domain Proteins using Transformers

MotivationBeing able to artificially design novel proteins of desired function is pivotal in many biological and biomedical applications. Generative statistical modeling has recently emerged as a new paradigm for designing amino acid sequences, including in particular models and embedding methods borrowed from Natural Language Processing (NLP). However, most approaches target single proteins or protein domains, and do not take into account any functional specificity or interaction with the context. To extend beyond current computational strategies, we develop a method for generating protein domain sequences intended to interact with another protein domain. Using data from natural multi-domain proteins, we cast the problem as a translation problem from a given interactor domain to the new domain to be generated, i.e. we generate artificial partner sequences conditional on an input sequence. ResultsEvaluating our models quality using diverse metrics, in part related to distinct biological questions, we show that our method outperforms state-of-the-art shallow auto-regressive strategies. We also explore the possibility of fine-tuning pre-trained large language models for the same task and of using Alphafold 2 for assessing the quality of sampled sequences.

bioinformatics↗

bioRxiv2022

DCAlign v1.0: Aligning biological sequences using co-evolution models and informative priors

DCAlign is a new alignment method able to cope with the conservation and the co-evolution signals that characterize the columns of a multiple sequence alignment of homologous sequences. However, the pre-processing steps required to align a candidate sequence are computationally demanding. We show in v1.0 how to dramatically reduce the overall computing time by including an empirical prior over an informative set of variables mirroring the presence of insertions and deletions. Availability and implementationDCAlign v1.0 is implemented in Julia and it is fully available at https://github.com/infernet-h2020/DCAlign Contactanna.muntoni@polito.it

bioinformatics↗

bioRxiv2021

AMaLa: analysis of Directed Evolution experiments via Annealed Mutational approximated Landscape

We present Annealed Mutational approximated landscape (AMaLa), a new method to infer fitness landscapes from Directed Evolution experiment sequencing data. Directed Evolution experiments typically start from a single wild-type sequence, which undergoes Darwinian in vitro evolution acted via multiple rounds of mutation and selection with respect to a target phenotype. In the last years, Directed Evolution is emerging as a powerful instrument to probe fitness landscapes under controlled experimental condition and, thanks to the use of high-throughput sequencing of the different rounds, as a relevant testing ground to develop accurate statistical models and inference algorithms. Fitness landscape modeling strategies, either use as input data the enrichment of variants abundances and hence require observing the same variants at different rounds, or they simply assume that the variants at the last sequenced round are the results of a sampling process at equilibrium. AMaLa aims at leveraging effectively the information encoded in the time evolution of all sequenced rounds. To do so, on the one hand we assume statistical sampling independence between sequenced rounds, and on the other we gauge all possible trajectories in sequence space with a time-dependent statistical weight consisting of two contributions: (i) a statistical energy term accounting for the selection process, (ii) a simple generalized Jukes-Cantor model to describe the purely mutational step. This simple scheme allows us to accurately describe the Directed Evolution dynamics in a concrete experimental setup and to infer a fitness landscape that reproduces correctly the measures of the phenotype under selection (e.g. antibiotic drug resistance), notably outperforming widely used inference strategies. We assess the reliability of AMaLa by showing how the inferred statistical model could be used to predict relevant structural properties of the wild-type sequence, and to reproduce the mutational effects of large scale functional screening not used to train the model.

biophysics↗

bioRxiv2021

Efficient generative modeling of protein sequences using simple autoregressive models

Generative models emerge as promising candidates for novel sequence-data driven approaches to protein design, and for the extraction of structural and functional information about proteins deeply hidden in rapidly growing sequence databases. Here we propose simple autoregressive models as highly accurate but computationally extremely efficient generative sequence models. We show that they perform similarly to existing approaches based on Boltzmann machines or deep generative models, but at a substantially lower computational cost. Furthermore, the simple structure of our models has distinctive mathematical advantages, which translate into an improved applicability in sequence generation and evaluation. Using these models, we can easily estimate both the model probability of a given sequence, and the size of the functional sequence space related to a specific protein family. In the case of response regulators, we find a huge number of ca. 1068 sequences, which nevertheless constitute only the astronomically small fraction 10-80 of all amino-acid sequences of the same length. These findings illustrate the potential and the difficulty in exploring sequence space via generative sequence models.

bioinformatics↗

bioRxiv2020

Aligning biological sequences by exploiting residue conservation and coevolution

Aligning biological sequences belongs to the most important problems in computational sequence analysis; it allows for detecting evolutionary relationships between sequences and for predicting biomolecular structure and function. Typically this is addressed through profile models, which capture position-specificities like conservation in sequences, but assume an independent evolution of different positions. RNA sequences are an exception where the coevolution of paired bases in the secondary structure is taken into account. Over the last years, it has been well established that coevolution is essential also in proteins for maintaining three-dimensional structure and function; modeling approaches based on inverse statistical physics can catch the coevolution signal and are now widely used in predicting protein structure, protein-protein interactions, and mutational landscapes. Here, we present DCAlign, an efficient approach based on an approximate message-passing strategy, which is able to overcome the limitations of profile models, to include general second-order interactions among positions and to be therefore universally applicable to protein- and RNA-sequence alignment. The potential of our algorithm is carefully explored using well-controlled simulated data, as well as real protein and RNA sequences.

bioinformatics↗

bioRxiv2020

Unsupervised inference of protein fitness landscape from deep mutational scan

The recent technological advances underlying the screening of large combinatorial libraries in high-throughput mutational scans, deepen our understanding of adaptive protein evolution and boost its applications in protein design. Nevertheless, the large number of possible genotypes requires suitable computational methods for data analysis, the prediction of mutational effects and the generation of optimized sequences. We describe a computational method that, trained on sequencing samples from multiple rounds of a screening experiment, provides a model of the genotype-fitness relationship. We tested the method on five large-scale mutational scans, yielding accurate predictions of the mutational effects on fitness. The inferred fitness landscape is robust to experimental and sampling noise and exhibits high generalization power in terms of broader sequence space exploration and higher fitness variant predictions. We investigate the role of epistasis and show that the inferred model provides structural information about the 3D contacts in the molecular fold.

bioinformatics↗