RESEARCH


Bioinformatics and Data Integration

Recent advances in science and technology are leading to a revision and reorientation of methodologies, addressing old and current issues from a new perspective. Advances in next-generation sequencing (NGS) enable comparative analysis of the abundance and diversity of whole microbial communities, generating large amounts of data and findings at the systems level. A current limitation for biologists is the growing demand for computational power and training to process NGS data. 

I am developing new Bioinformatics tools and designing new scientific workflows to integrate multi-omics data (for example, metagenomics and metatranscriptomics datasets).

(A) Bacteriophage data representation is based on DNA sequencing approaches, followed by the assembly of short reads. (B) Model architecture for deep learning on metagenomic data to predict bacteriophage genomic sequencing from metagenomic data. [Silva et al., 2025; doi: 10.1016/j.virol.2025.110559]
(A) Bacteriophage data representation is based on DNA sequencing approaches, followed by the assembly of short reads. (B) Model architecture for deep learning on metagenomic data to predict bacteriophage genomic sequencing from metagenomic data. [Silva et al., 2025; doi: 10.1016/j.virol.2025.110559]

One Health Relationships between Human, Animal, and Environmental Microbiomes

The One Health concept is a global strategy to study the relationship between human and animal health and the transfer of pathogenic and non-pathogenic species between these systems. In clinical settings, metagenomics is also a powerful tool in the fight against antibiotic-resistant pathogens in humans and animals. 

Currently, I am further exploring the role of antibiotic-resistance genes in microbial communities across several ecosystems, including host-associated and soil microbiomes, in a One Health context.

Reconstruction of ~ 3,000 microbial genomes from a large-scale sampling of animals using One Health concept. (Lemos et al. 2022 - [Lemos et al. 2022 - Scientific Data] doi: 10.1038/s41597-022-01465-5)
Reconstruction of ~ 3,000 microbial genomes from a large-scale sampling of animals using One Health concept. (Lemos et al. 2022 - [Lemos et al. 2022 - Scientific Data] doi: 10.1038/s41597-022-01465-5)



Genome-Resolved Metagenomics

Despite efforts to access microbial diversity, most soil microbes remain unknown, and we are still far from understanding several microbe-mediated processes in soil. The development and application of novel computational strategies have successfully allowed us to reconstruct complete or near-complete genomes of rare and/or uncultured bacteria.

I am working on the application of computational methods to reconstruct microbial genomes from metagenomic data using several types of datasets (from low-diversity to highly complex environments).


Lemos et al. (2021) Trends in Microbiology - [https://www.sciencedirect.com/science/article/pii/S0966842X21000159]
Lemos et al. (2021) Trends in Microbiology - [https://www.sciencedirect.com/science/article/pii/S0966842X21000159]

Computational identification of CRISPR-Cas system-associated genes in the intestinal microbiota of obese individuals

Intestinal dysbiosis is considered a factor contributing to complex diseases, including human obesity. Recent high-throughput DNA sequencing studies indicate that obese patients exhibit reduced microbial diversity. However, many functional aspects of microbial communities remain underexplored, including the diversity of genes associated with the CRISPR-Cas system — a microbial defense mechanism against bacteriophages that stores information from past infections. In this context, the main objective of this project is to identify and quantify CRISPR-associated genes in the gut microbiota of obese individuals and compare them with those in non-obese individuals. The analysis will focus on in silico comparative analyses using public databases of genetic sequences and genomes reconstructed from metagenomes. We aim to determine CRISPR array orientation, annotate Cas proteins, and classify their subtypes. This will allow us to determine whether the microbiota of obese individuals exhibits higher or lower CRISPR diversity than that of non-obese individuals, and to highlight potential new genome-editing biotechnologies. This project will also contribute to training human resources in Data Science, Microbiology, and Bioinformatics within the interdisciplinary context of Ilum (School of Science) at the Center for Research in Energy and Materials.

Collaborative Research

None of what I am doing would be possible without the collaboration and help of all the people who composed the multidisciplinary teams that I have worked with during my scientific career.