Peptide-based drug discovery through artificial intelligence: towards an autonomous design of therapeutic peptides.

With their diverse biological activities, peptides are promising candidates for therapeutic applications, showing antimicrobial, antitumour and hormonal signalling capabilities. Despite their advantages, therapeutic peptides face challenges such as short half-life, limited oral bioavailability and susceptibility to plasma degradation. The rise of computational tools and artificial intelligence (AI) in peptide research has spurred the development of advanced methodologies and databases that are pivotal in the exploration of these complex macromolecules. This perspective delves into integrating AI in peptide development, encompassing classifier methods, predictive systems and the avant-garde design facilitated by deep-generative models like generative adversarial networks and variational autoencoders. There are still challenges, such as the need for processing optimization and careful validation of predictive models. This work outlines traditional strategies for machine learning model construction and training techniques and proposes a comprehensive AI-assisted peptide design and validation pipeline. The evolving landscape of peptide design using AI is emphasized, showcasing the practicality of these methods in expediting the development and discovery of novel peptides within the context of peptide-based drug discovery.

Autonomous B-cell receptor signaling and genetic aberrations in chronic lymphocytic leukemia-phenotype monoclonal B lymphocytosis in siblings of patients with chronic lymphocytic leukemia.

Clonal expansion of CD5-expressing B cells, commonly designated as monoclonal B lymphocytosis (MBL), is a precursor condition for chronic lymphocytic leukemia (CLL). The mechanisms driving subclinical MBL B-cell expansion and progression to CLL, occurring in approximately 1% of affected individuals, are unknown. An autonomously signaling B-cell receptor (BCR) is essential for the pathogenesis of CLL. The objectives of this study were functional characterization of the BCR of MBL in siblings of CLL patients and a comparison of genetic variants in MBL-CLL sibling pairs. Screening of peripheral blood by flow cytometry detected 0.2-480 clonal CLL-phenotype cells per microliter (median: 37/µL) in 34 of 191 (17.8%) siblings of CLL patients. Clonal BCR isolated from highly purified CLL-phenotype cells induced robust calcium mobilization in BCR-deficient murine pre-B cells in the absence of external antigen and without experimental crosslinking. This autonomous BCR signal was less intense than the signal originating from the CLL BCR of their CLL siblings. According to genotyping by single nucleotide polymorphism array, whole exome, and targeted panel sequencing, CLL risk alleles were found with high and similar prevalence in CLL patients and MBL siblings, respectively. Likewise, the prevalence of recurrent CLL-associated genetic variants was similar between CLL and matched MBL samples. However, copy number variations and small variants were frequently subclonal in MBL cells, suggesting their acquisition during subclinical clonal expansion. These findings support a stepwise model of CLL pathogenesis, in which autonomous BCR signaling leads to a non-malignant (oligo)clonal expansion of CD5+ B cells, followed by malignant progression to CLL after acquisition of pathogenic genetic variants.

Integration of Mutational Signature Analysis with 3D Chromatin Data Unveils Differential AID-Related Mutagenesis in Indolent Lymphomas.

Activation-induced deaminase (AID) is required for somatic hypermutation in immunoglobulin genes, but also induces off-target mutations. Follicular lymphoma (FL) and chronic lymphocytic leukemia (CLL), the most frequent types of indolent B-cell tumors, are exposed to AID activity during lymphomagenesis. We designed a workflow integrating de novo mutational signatures extraction and fitting of COSMIC (Catalogue Of Somatic Mutations In Cancer) signatures, with tridimensional chromatin conformation data (Hi-C). We applied the workflow to exome sequencing data from lymphoma samples. In 33 FL and 30 CLL samples, 42% and 34% of the contextual mutations could be traced to a known AID motif. We demonstrate that both CLL and FL share mutational processes dominated by spontaneous deamination, failures in DNA repair, and AID activity. The processes had equiproportional distribution across active and nonactive chromatin compartments in CLL. In contrast, canonical AID activity and failures in DNA repair pathways in FL were significantly higher within the active chromatin compartment. Analysis of DNA repair genes revealed a higher prevalence of base excision repair gene mutations (p = 0.02) in FL than CLL. These data indicate that AID activity drives the genetic landscapes of FL and CLL. However, the final result of AID-induced mutagenesis differs between these lymphomas depending on chromatin compartmentalization and mutations in DNA repair pathways.

The role of machine learning in health policies during the COVID-19 pandemic and in long COVID management.

The ongoing COVID-19 pandemic is arguably one of the most challenging health crises in modern times. The development of effective strategies to control the spread of SARS-CoV-2 were major goals for governments and policy makers. Mathematical modeling and machine learning emerged as potent tools to guide and optimize the different control measures. This review briefly summarizes the SARS-CoV-2 pandemic evolution during the first 3 years. It details the main public health challenges focusing on the contribution of mathematical modeling to design and guide government action plans and spread mitigation interventions of SARS-CoV-2. Next describes the application of machine learning methods in a series of study cases, including COVID-19 clinical diagnosis, the analysis of epidemiological variables, and drug discovery by protein engineering techniques. Lastly, it explores the use of machine learning tools for investigating long COVID, by identifying patterns and relationships of symptoms, predicting risk indicators, and enabling early evaluation of COVID-19 sequelae.

Mutation in a SARS-CoV-2 Haplotype from Sub-Antarctic Chile Reveals New Insights into the Spike's Dynamics.

The emergence of SARS-CoV-2 variants, as observed with the D614G spike protein mutant and, more recently, with B.1.1.7 (501Y.V1), B.1.351 (501Y.V2) and B.1.1.28.1 (P.1) lineages, represent a continuous threat and might lead to strains of higher infectivity and/or virulence. We report on the occurrence of a SARS-CoV-2 haplotype with nine mutations including D614G/T307I double-mutation of the spike. This variant expanded and completely replaced previous lineages within a short period in the subantarctic Magallanes Region, southern Chile. The rapid lineage shift was accompanied by a significant increase of cases, resulting in one of the highest incidence rates worldwide. Comparative coarse-grained molecular dynamic simulations indicated that T307I and D614G belong to a previously unrecognized dynamic domain, interfering with the mobility of the receptor binding domain of the spike. The T307I mutation showed a synergistic effect with the D614G. Continuous surveillance of new mutations and molecular analyses of such variations are important tools to understand the molecular mechanisms defining infectivity and virulence of current and future SARS-CoV-2 strains.