Services

Professional Bioinformatics Services

The Structural Bioinformatics and Computational Biology Laboratory (SBCB Lab) at the Federal University of Rio Grande do Sul (UFRGS) offers bioinformatics analysis, consulting and training to academic, public and private research groups. Our services draw on the methods the laboratory develops and publishes in its own research: statistical and machine-learning analysis of omics data, heuristic optimization, and structural modeling. Every analysis is documented and reproducible, and the results are delivered with an explicit description of their assumptions and limitations.

• Bioinformatics analysis: end-to-end processing and analysis of your data, from raw files and quality control to statistical results, figures and a methods report.
• Bioinformatics consulting: support with experimental design, choice of methods and interpretation, for teams that run their own analyses.
• Bioinformatics training: lectures and hands-on courses tailored to your team’s level and data types.

Contracting Information: Services are contracted formally through the Luiz Englert Foundation (FLE), the support foundation that administers the agreement (CNPJ 92.971.845/0001-42).

We analyse high-throughput biological data using established, peer-reviewed pipelines, and custom methods where needed. Typical projects include:

• Machine learning and artificial intelligence: supervised and unsupervised machine-learning and deep-learning models for classification, prediction, feature (biomarker) selection and data integration, including model interpretability analyses. Models are evaluated with nested or held-out validation, appropriate performance metrics and, when available, independent datasets, so that performance estimates are not optimistically biased.

• Genomics: quality control and read processing; de novo and reference-guided genome assembly and annotation; variant calling and annotation (SNVs, small indels, structural variants/CNVs); pangenome and comparative genomics; genetic ancestry and population-structure analysis; pharmacogenomic variant interpretation; and phylogenetic and evolutionary analyses.

• Transcriptomics: bulk RNA-seq, single-cell RNA-seq and microarray analysis - quality control, normalization and batch-effect assessment, differential expression with multiple-testing correction, functional enrichment (GO, pathways), co-expression networks and, for single-cell data, clustering, cell-type annotation and trajectory analysis. Long non-coding RNA analyses are also available.

• Microbiome - amplicon and shotgun metagenomics: Amplicon sequencing (16S rRNA, ITS, 18S) - ASV inference, taxonomic assignment, alpha and beta diversity, and differential-abundance analysis with compositional-data-aware methods; Shotgun metagenomics - read-based taxonomic and functional profiling, assembly and binning into metagenome-assembled genomes (MAGs), and detection of antimicrobial-resistance genes (resistome) and mobile genetic elements; Longitudinal or multi-site designs: - analyses of community dynamics over time or across environments.

• Custom methods and pipelines: development of analysis pipelines or new methods when standard tools are not adequate - for example, optimization with heuristic search methods, feature-selection strategies, molecular docking and molecular dynamics analyses, or GPU-accelerated implementations.

Consulting is intended for teams that will run their own analyses but want methodological support at specific stages of a project.

• Experimental design: definition of the biological question and comparisons, sample-size and statistical-power estimation, replication strategy (biological vs. technical replicates), randomization and control of batch effects, and choice of sequencing strategy and depth, before samples are collected or sequenced.

• Analytical strategy and tool selection: choosing methods, reference databases and parameters appropriate to the data type and question, with justification based on published benchmarks where they exist.

• Interpretation of results: placing statistical results in biological context, distinguishing association from causation, assessing effect sizes and uncertainty, and identifying which findings require experimental validation.

• Workflow review and troubleshooting: review of existing pipelines, code and results to identify technical or methodological problems (e.g., inadequate normalization, data leakage in machine-learning models, incorrect statistical tests).

• Multi-omics integration: selection of integration strategies (early, intermediate or late integration; matrix-factorization, network-based and deep-learning approaches) according to study design, sample overlap between omics layers and sample size.

We offer lectures, short courses and hands-on workshops, delivered on site or online and tailored to the audience:

• Introductory courses: biological databases, file formats (FASTA, FASTQ, SAM/BAM, VCF), command-line and Python/R basics, and core analysis tools, with hands-on exercises.

• Advanced modules: complete workflows for a specific data type (e.g., RNA-seq, single-cell, metagenomics, variant analysis, molecular docking and dynamics, machine learning for omics), from raw data to interpretation, including good practices for reproducibility (version control, workflow managers, containers).

Veterinary microbial genomics (in partnership with the Laboratório de Bacteriologia Veterinária - UFRGS)

• Genomic detection of antimicrobial-resistance determinants: identification of resistance genes, mutations and associated mobile genetic elements in bacterial isolates or samples from swine, cattle and poultry. Genotypic results complement, but do not replace, phenotypic antimicrobial susceptibility testing performed according to recognized standards.

• Bacterial isolate characterization: whole-genome-based species identification, typing (e.g., MLST/cgMLST), virulence-gene profiling and phylogenetic comparison of isolates, supporting epidemiological investigation.

• Deliverables: methods report (software versions, parameters, databases), result tables and figures, and analysis code or workflow.
• Reproducibility: version-controlled code and documented software environments.
• Data confidentiality: restricted-access storage; data used only for the contracted project; LGPD compliance for personal or health data.

To discuss a project, send a short description of your question, data type, number of samples and timeline to mdorn@inf.ufrgs.br. We will reply with a proposed scope and next steps.

Location: Center for Biotechnology (CBiot) and Institute of Informatics (INF), Federal University of Rio Grande do Sul (UFRGS), Av. Bento Gonçalves 9500, Porto Alegre, RS, Brazil.

Analyses are run on the laboratory’s own computing infrastructure: GPU-equipped servers (including NVIDIA RTX PRO 6000 Blackwell and RTX A6000 accelerators), up to 1 TB of RAM per node and more than 900 TB of storage. See more details here.