New tool could facilitate search for disease-causing genetic variants

Methodology is being developed by multinational Seven Bridges in collaboration with the Brazilian Initiative on Precision Medicine. Novelty was presented during the 4th BRAINN Congress.

By Karina Toledo, in Campinas  |  Agência FAPESP – Scientists affiliated with the Brazilian Initiative on Precision Medicine (BIPMed) are working in partnership with Seven Bridges, a multinational that specializes in biomedical data analysis, to build a new tool that could facilitate the identification of disease-causing genetic variants.

The novelty was presented at a workshop held at the University of Campinas (UNICAMP) in São Paulo State on March 25 as part of the 4th BRAINN Congress organized by the Brazilian Research Institute for Neuroscience and Neurotechnology (BRAINN), one of the Research, Innovation and Dissemination Centers (RIDCs) funded by FAPESP.

Known as The Graph Genome Suite, the methodology that is being developed creates reference genomes in the form of graphs using genetic data for an entire population.

“At present, when you analyze DNA fragments, the sequencer gives you a jigsaw puzzle with millions of pieces that must be put in the right order. To do that, you use a reference genome. The trouble is that the reference genome, the map we use as a guide, derives from data for only a few individuals and doesn’t reflect the genetic complexity of the Brazilian population,” said Benilton Carvalho, a researcher at UNICAMP’s Mathematics, Statistics & Scientific Computation Institute (IMECC) and a member of BIPMed’s steering committee.

The explanation was taken further by Iscia Lopes-Cendes, a professor at UNICAMP’s Medical School (FCM) and a member of BRAINN and BIPMed, who explained that the human genome is highly variable. “You can find huge pieces that are duplicated or lacking when you compare different individuals,” she said. “That isn’t necessarily a problem. It’s just the normal variation.”

However, when researchers in the laboratory come across a fragment not found in the reference genome, they tend to ignore it. “That poses a problem if you’re looking for a disease-causing genetic variant and it happens to be in the piece you’ve discarded,” Cendes said.

According to Carvalho, the new tool will not assemble the puzzle in linear fashion, but rather in the form of a diagram or graph, so that any apparently mismatched pieces will be added as alternatives to the main path instead of being ignored. This method will create slightly different versions of the same reference genome.

“We’re working with Seven Bridges on two fronts. One is development of the methodology itself, and the other entails supplying information on the differences we observe in the Brazilian population compared with the reference genome. This participation is crucial so that, when the tool is ready, it can also be used for research done in Brazil,” said Cendes, one of the founders of BIPMed.

Genomic Google

Another innovation presented by BIPMed’s bioinformatics team at the workshop held in Campinas is a search engine for genomics databases.

“Today, when users access BIPMed’s database, for example, they can look up one or two variants, maybe ten. But querying 20,000, for example, would take such a long time that it wouldn’t be feasible. The new tool will help with that kind of programmatic search. It’s a sort of Google for genomics databases,” Carvalho said.

The software is in the public domain and is now available from Bioconductor, a community created for the sharing of open-source bioinformatics programs.

“It’s important to bear in mind that this tool isn’t for use only with BIPMed’s database. It’s also designed to connect to any federated database that uses the standards established by the Global Alliance for Genomics and Health (GA4GH). All you have to do is enter the server’s address,” Carvalho said.

BIPMed’s genomics database is the first initiative of its kind in Latin America. It already contains data for 254 individuals from the Campinas region, volunteers (not patients) representing a reference population whose complete exome has been sequenced or whose genetic material has been analyzed by microarray. Whole-exome sequencing (WES) is a technique for identifying disease-causing DNA variants in the genome’s exons, the pieces of genes that code for proteins.

According to Cendes, the database has also recently begun to include information on patients with epilepsy who receive care from UNICAMP’s health service. “Seventeen different mutations of SCN1A, one of the genes most closely associated with epilepsy, have been registered, and the number is expected to grow rapidly,” she said.

Data for patients with syndromic or non-syndromic cleft lip and palate and for patients with breast cancer-related mutations will soon be included as well.

Nationwide scope

To establish the conditions necessary for precision medicine to be implemented in Brazil, BIPMed was founded in 2015 by researchers affiliated with BRAINN and four other RIDCs funded by FAPESP: the Center for Computational Engineering and Sciences (CCES), the Center for Cell-Based Therapy (CTC), the Obesity and Comorbidities Research Center (OCRC), and the Center for Research on Inflammatory Diseases (CRID). 

BIPMed has won international recognition as a pioneering initiative and now represents Brazil in important consortia such as GA4GH and The Human Variome Project, an NGO linked to the United Nations Educational, Scientific & Cultural Organization (UNESCO).

“At the end of 2016, we attended a meeting of national initiatives in genomic medicine organized by GA4GH in Vancouver, Canada. The idea was that each group would present on the work it’s been doing individually, to foster an exchange of experiences and mutual assistance,” Cendes said.

Also present were representatives of the Australian Genomics Health Alliance, the France Genomic Medicine Plan 2025, Genomics England, Genome Canada, H3Africa, the Japan Agency for Medical Research & Development, the US Precision Medicine Initiative, and the US Cancer Moonshot.

“All the groups represented at the meeting were linked to the health systems of their respective countries of origin. In many cases, it was the public health system,” Cendes said. “BIPMed was the only initiative backed by a research funding agency. This will have to change over time because, at some stage, we must stop producing scientific articles and start delivering healthcare services. And then we’re going to need other sources of funding.”

In her view, the managers of Brazil’s Unified Health System (SUS) are too reluctant to recognize genomic medicine as a discipline that the public health system can use or even that has any practical utility at all. This attitude is behind the times, according to Cendes.

“You often hear it said that genomic data are useful only to diagnose and treat rare diseases,” she said. “And yet at cutting-edge private centers, genomic medicine is already being used for staging, grading and defining common diseases, such as cancer, which is important to ensure patients don’t receive more aggressive treatment than necessary and to save money.”

For Peter Goodhand, the executive director of GA4GH, BIPMed has taken a major step toward bringing “the power and promise of genomic medicine” to Brazil and all of Latin America. 

“It’s important that it continues to grow in terms of participation and outreach,” he added. “That will require strong leadership and funding from multiple partners at both the regional and national levels.”