Nicola Senin

Antimicrobial resistance (AMR) is projected to increase globally, yet the genomic traits and social determinants of health driving its trajectory are poorly understood. Here, we present the first global-scale AMR forecasting analysis, integrating machine learning, Monte Carlo simulations and forecasting modelling to identify clinically relevant AMR traits projected to increase, and the key determinants driving their increase over the next 25 years. We analysed 45,645 bacterial genomes (including ESKAPE and World Health Organisation high/critical priority pathogens), 298,288 resistance profiles, spanning 84 antibiotics across 27 antimicrobial classes and 1112 determinants of health across 125 countries. We identified 78 clinically relevant AMR feature-pathogen pairs projected to rise globally by 2050. Among these, 94.9% were multidrug resistant (MDR), 89.7% globally widespread, 80.8% persisting over time, and 48.7% multi-host. Our analysis reveals mortality indicators (68.79%) are the strongest predictors, followed by socioeconomic factors (15.29%), and antibiotic consumption (4.46%). However, at the molecular level, socioeconomic disparities are the primary drivers of AMR, strongly associated to the fourteen most critical features (aadA, aadA2, APH(3')-Ia, APH(3'')-Ib, APH(6)-Id, ermB, mphA, qacEdelta1, rep_cluster_1118, sul1, sul2, TEM-34, tet(A), and tet(M)). These features emerged as primary health threats, clinically relevant, MDR, human-enriched, inter-species spreaders, plasmid-associated, and present across up to 10 hosts in at least four World Bank regions. By understanding the interplay of biological and social factors in shaping AMR trends to 2050, we provide a roadmap for targeted AMR mitigation.

In Bangladesh, Vibrio cholerae lineages are undergoing genomic evolution, with increased virulence and spreading ability. However, our understanding of the genomic determinants influencing lineage transmission and disease severity remains incomplete. Here, we developed a computational framework using machine-learning, genome scale metabolic modelling (GSSM) and 3D structural analysis, to identify V. cholerae genomic traits linked to lineage transmission and disease severity. We analysed in-patients isolates from six Bangladeshi regions (2015-2021), and uncovered accessory genes and core SNPs unique to the most recent dominant lineage, with virulence, motility and bacteriophage resistance functions. We also found a strong correlation between V. cholerae genomic traits and disease severity, with some traits overlapping those driving lineage transmission. GSMM and 3D structure analysis unveiled a complex interplay between transcription regulation, protein interaction and stability, and metabolic networks, associated to lifestyle adaptation, intestinal colonization, acid tolerance and symptom severity. Our findings support advancing therapeutics and targeted interventions to mitigate cholera spread.

Sharing of genetic elements among different pathogens and commensals inhabiting same hosts and environments has significant implications for antimicrobial resistance (AMR), especially in settings with high antimicrobial exposure. We analysed 661 Escherichia coli and Salmonella enterica isolates collected within and across hosts and environments, in 10 Chinese chicken farms over 2.5 years using data-mining methods. Most isolates within same hosts possessed the same clinically relevant AMR-carrying mobile genetic elements (plasmids: 70.6%, transposons: 78%), which also showed recent common evolution. Supervised machine learning classifiers revealed known and novel AMR-associated mutations and genes underlying resistance to 28 antimicrobials, primarily associated with resistance in E. coli and susceptibility in S. enterica. Many were essential and affected same metabolic processes in both species, albeit with varying degrees of phylogenetic penetration. Multi-modal strategies are crucial to investigate the interplay of mobilome, resistance and metabolism in cohabiting bacteria, especially in ecological settings where community-driven resistance selection occurs.

China is the largest global consumer of antimicrobials and improving surveillance methods could help to reduce antimicrobial resistance (AMR) spread. Here we report the surveillance of ten large-scale chicken farms and four connected abattoirs in three Chinese provinces over 2.5 years. Using a data mining approach based on machine learning, we analysed 461 microbiomes from birds, carcasses and environments, identifying 145 potentially mobile antibiotic resistance genes (ARGs) shared between chickens and environments across all farms. A core set of 233 ARGs and 186 microbial species extracted from the chicken gut microbiome correlated with the AMR profiles ofEscherichia colicolonizing the same gut, includingArcobacter, AcinetobacterandSphingobacterium, clinically relevant for humans, and 38 clinically relevant ARGs. Temperature and humidity in the barns were also correlated with ARG presence. We reveal an intricate network of correlations between environments, microbial communities and AMR, suggesting multiple routes to improving AMR surveillance in livestock production.

Quality monitoring for laser powder bed fusion (L-PBF), particularly in-process and real-time monitoring, is of importance for part quality assurance and manufacturing cost reduction. Measurement of layer surface topography is critical for quality monitoring, as any anomaly on layer surfaces can result in defects in the final part. In this paper, we propose a surface measurement method, based on the use of scattered light patterns and a convolutional autoencoder-based unsupervised machine learning method, designed and trained using a large set of scattering patterns simulated from reference surfaces using a scattering model. The advantage of using an autoencoder is that the monitoring model can be trained using solely data from acceptable surfaces, without the need to ensure the presence of representative observations for all the types of possible surface defects. The advantage of using simulated data for training is that we can obtain an effective monitoring solution without the need for a large collection of experimental observations. Here we report the results of a preliminary investigation on the performance of the proposed solution, where the trained autoencoder is tested on experimental data obtained off-process, using a dedicated experimental apparatus for generating and collecting light scattering patterns from manufactured L-PBF surfaces. Our results indicate that the proposed monitoring solution is capable of detecting both acceptable and anomalous surfaces. Although further validation is required to fully assess performance within an on-machine and in-process setup, our preliminary results are encouraging and provide a glimpse of the potential benefits of using our surface measurement solution for L-PBF in-process monitoring.

This paper presents a novel quality monitoring method for additive manufactured surfaces combining machine learning and light scattering. The proposed method aims to monitor undesired topographical modifications of additive manufactured surfaces by detecting changes in a scattering pattern using an autoencoder, which is an unsupervised machine learning model, trained with datasets directly measured from reference surfaces with desired surface topographies. Given the unsupervised learning nature of the autoencoder, training with datasets acquired from surfaces with deviations is not necessary, which makes the proposed method appealing, as there is no need to retrieve defective surface samples to train the autoencoder. More importantly, the autoencoder can be updated when datasets from a new type of surface with desired but different topographies are available. As scattering patterns related to new topographies are relatively easy to obtain by experiment, we demonstrate that our autoencoder can be retrained with new scattering patterns and learn to address a wider variety of surfaces, showing superior performance with respect to machine learning solutions adopting a static model, trained only once on the initially available information. Experiments performed on laser powder bed fusion surfaces show that the proposed method is effective. The relatively simple and low-cost setup of the measurement system also makes the proposed method appealing for implementation on commercial additive manufacturing machines.