Stephen Hart, PhD, a Senior Research Scientist at Frontier Science Foundation (FSF), is the lead author of an article published in the September 1, 2018 edition of the Journal of Acquired Immune Deficiency Syndromes (JAIDS), “Impact of changes over time in the Stanford University genotypic resistance algorithm.” The article comes out of ACTG Data Analysis Concept Sheet (DACS) 330, which Steve designed and led. Sarah Strobino, BS, also from Frontier Science Foundation, made major contributions to the project. Saran Vardhanabhuti, PhD and Linda J. Harrison, MSc, from Harvard T.H. Chan School of Public Health, did the trend analysis.
The Stanford algorithm is the most widely used method of predicting resistance to anti-retroviral drugs from genetic sequences or mutation lists for the HIV-1 found in a patient’s body. It provides rules for deciding whether a patient has no resistance, high-level resistance, or one of three intermediate levels, based on the mutations that have evolved in the patient’s population of HIV-1, mutations that can make the population less susceptible to the effects of one or more drugs.
The article looks at differences in drug resistance predictions, based on the same specimens and genetic sequences, depending on the version of the Stanford algorithm that is used. For most drugs, more recent versions result in a prediction of higher drug resistance, even when the specimens are exactly the same. The problem this causes is that researchers often compare their results to results from other studies, sometimes studies in other countries, or a few years earlier in the same country. Usually the algorithm versions used in the other studies are unknown, and in all probability different, making the comparisons uninformative. This article concludes that algorithm version matters and close attention to its impact is essential for accumulating and comparing HIV-1 drug resistance results across different populations and over time.
Technology developed by Frontier Science, the Algorithm Specification Interface (ASI) Interpreter, made this research possible. ASI Interpreter, an open source Java library developed by FSF programmers, was used to make antiretroviral drug resistance calculations. The ASI Interpreter library allows any version of any algorithm, provided that it is described in an XML file, to be compiled and executed, on the fly, on any set of sequences. For DACS 330, 14 algorithms were each executed on sequences from more than 5,000 patients. The ASI Interpreter library has now been released as open-source software, on GitHub, for the benefit of the research community.
The citation for the article is:
Impact of Changes Over Time in the Stanford University Genotypic Resistance Interpretation Algorithm. Hart, Stephen A., PhD; Vardhanabhuti, Saran, PhD; Strobino, Sarah A., BS; Harrison, Linda J., MSc. JAIDS Journal of Acquired Immune Deficiency Syndrome: Sep. 1, 2018; Vol. 79, Issue 1; p e21–29
If you wish to read the article, e-mail Steve at hart@fstrf.org