Software enables COVID vaccines

Picture of ANU School of Computing

Written by: ANU School of Computing
1 Sep 2021


Dr. Minh Bui
Dr. Minh Bui

He is an indispensable leader in humanity’s war against the coronavirus. But outside the highly technical fields of biomedical research and bioinformatics software, few people know of Dr Minh Bui, a senior lecturer at the College of Engineering & Computer Science at The Australian National University (ANU).

Bui’s 2014 invention IQ-TREE is an open source software tool that has helped scientists around the globe synthesize enormous volumes of genetic data to map the evolutionary trees of biological entities—a process called phylogenomic inference. Bui’s software has provided crucial information for the development of vaccines and for contact tracing. IQ-TREE also played a key role in identifying the Delta Variant many months before it reached Australian shores.

Today, the Chan Zuckerberg Initiative (CZI) announced that Bui and his research team have been awarded grant funding for the second time in three years. The new grant is for US $340,000, bringing the total to nearly $464,000 as part of CZI’s Essential Open Source Software for Science (EOSS) program.

Spurred by CZI’s 2019 grant, Bui and his team were only a few months into the project of improving the speed and scalability of IQ-TREE when the pandemic hit in 2020. Progress was hindered as Australia entered a hard lockdown. The team was unable to organise in-person gatherings to test iterations of the software.

“However, it also impacted our work in a positive way, because IQ-TREE was being used by researchers to study the evolution of this virus,” Bui said.

The global scientific community had leapt headlong into a genomic sequencing effort on an unprecedented scale. IQ-TREE wasn’t perfect, but it was the best tool available to cope with the deluge of genetic data. Mindful of the software’s new importance, Bui sacrificed social and leisure activities to spend the majority of his time in the lab.

His team published IQ-TREE 2.0 in May of 2020. It could infer an evolutionary tree from 17,000 COVID sequences in only three minutes whereas version 1 had required 37 hours to do the same. It was a breakthrough. The two versions of the software were downloaded more than 50,000 times in 2020, and the genomic data they produced contributed directly the indentification of variants and the development of the COVID-19 vaccines.

In their grant proposal, the team argued that the continued development of IQ-TREE was urgently needed to hunt after new variants in the current pandemic, and also to meet the challenges of future biomedical crises.

Before the pandemic, Bui had dedicated much of his time to the IQ-TREE user community, answering questions at a rate of two enquiries per day. As the virus began seeding in populations around the world, the volume of questions increased to the point that Bui became overwhelmed. “Now as this grant is successful, I will hire someone who will help with this,” he said.

IQ-TREE’s effectiveness in tracking the COVID-19 virus and its mutations has been a prime example of CZI’s mission to cure disease via technology.

“Open source software for biomedical research is a critical component of an open, reproducible, and verifiable scientific ecosystem,” said CZI Head of Science Cori Bargmann.

Scientific open source development often lacks dedicated funding for updating and improvement. CZI is filling that void through EOSS, with the recently announced awards bringing its total funding commitment to over $28 million.

Bui first conceived of the idea for IQ-TREE in 2011 while working at Max Perutz Labs in Vienna, Austria. He was frustrated by the limitations of the state-of-the-art applications at the time. “Some applications were fast but not so accurate, and some were more accurate but slow. So my goal was to have something both fast and accurate,” he said.

IQ-TREE was then developed at the Center for Integrative Bioinformatics Vienna by Bui and colleagues Lam-Tung Nguyen, Heiko A. Schmidt, Arndt von Haeseler. In November of 2014, the team published a paper introducing IQ-TREE 1.0.

“I never imagined that it would become so popular,” Bui said. IQ-TREE quickly became one of the two most widely used tools for phylogenetic inference by maximum likelihood.

“I think the success was because IQ-TREE was not just designed as a scientific software. Instead we applied a lot of industrial standards in software development to make it very user friendly,” he said.

Associate Professor Robert Lanfear of the ANU School of Biology had been impressed by IQ-TREE 1.0, but beyond that it was “Minh’s expertise and enthusiasm for making phylogenetics better” that inspired Lanfear to recruit him to come to ANU.

Lanfear said he could not have foreseen the full significance of Bui’s work at the time. “But, it was already very clear that IQ-TREE was one of a very small number of leading pieces of software in the field.”

Bui said that the combination of international prestige and cross-disciplinary collaboration brought him from Vienna to Canberra. “I saw many scientists working in the relevant field of evolutionary biology and I wanted to broaden my research,” Bui said.

Bui accepted the ANU research fellowship in 2018 and, among other projects, began collaborating with Lanfear on IQ-TREE. “Many ideas for IQ-TREE came from my fruitful discussions with him,” Bui said.

Biu spent a year at the ANU School of Biology before being hired by the School of Computing in 2019. Lanfear said he was “very happy” to lose him.

“My funding for Minh’s position was fairly short term,” Lanfear explained. “Ultimately I just want the best for Minh, and his move to Computing represented a step up for his career in allowing him to start his own research group, and, it improved the chance that ANU will get to keep him for longer—hopefully indefinitely.”

In addition to Lanfear, Bui said the IQ-TREE project relies heavily upon School of Computing colleague Dr Giuseppe Barca, who can address high performance computing demands, and James Barbetti , a software engineer in Bui’s lab.

The new funding from CZI puts Bui and his team in position to enable real-time genomic epidemiology during ongoing outbreaks such as COVID-19 and develop IQ-TREE 3.0 complete with updates to address feedback from hundreds of scientists who contacted them during the pandemic.

The work will be jointly conducted between the ANU School of Computing and School of Biology, where Lanfear will be the main point of contact. Bui said the School of Biology has a very strong department in ecology and evolution with well-known academics such as Craig Moritz, Lindell Bromham, Gavin Huttley, Michael Jennions, and Scott Keogh.

“The idea is that computer scientists will develop and implement new models and methods while biologists cover data analysis,” Bui said, anticipating that Amanda Barnard and Yu Lin—of the newly established computational science cluster—will be invaluable contributors.

“In the next couple of years I think we’ll make big progress on ‘real time’ phylogenetics, where we continuously update phylogenetic trees for use in public health situations like SARS-CoV-2,” Lanfear said. “That work is thanks to the CZI funding.”

The team also expects to make progress on accounting for recombination and hybridisation thanks to a grant from the Australian Research Council (ARC), and on estimating better models of evolution that more accurately reflect reality. And Bui’s PhD student Nhan Trong Ly is extending IQ-TREE so that it can simulate datasets.

Bui believes his work has benefited tremendously from the interdisciplinary environment at the ANU. “Nowhere else would be comparable,” he said.

IQ-TREE researchers
Dr. Minh Bui meets with fellow IQ-TREE researchers James Barbetti (top left), Associate Professor Robert Lanfear (bottom left), and Dr Giuseppe Barca (bottom right)
arrow-left bars search times arrow-up