Clinical Microbiomics debuts “revolutionary” human microbiome profiling tool
21 Sep 2023 --- Denmark-based Clinical Microbiomics has developed the Human Microbiome Profiler 2.0 called Champ, which is set to accelerate microbiome research by maximizing insights from data and minimizing the risk of false discoveries.
Demonstrating high sensitivity and accuracy in microbiome profiling, Champ leverages Clinical Microbiomics’ proprietary algorithm calibrated across thousands of analyzed samples to ensure results are robust to the natural diversity across human microbiome samples.
“With Champ, we introduce our next-generation human microbiome profiler. It’s based on the world’s largest collection of high-quality human microbiome data, encompasses the full microbial diversity across all body sites, and assures profiling with an unmatched accuracy and incredibly low risk of false signals. Dr. H. Bjørn Nielsen, chief scientific officer of Clinical Microbiomics, tells Nutrition Insight.
“Correct profiling is the cornerstone of a sound microbiome study – without it, conclusions can easily be misguided.”
Champ was developed by Nielsen and his team of bioinformatics experts at Clinical Microbiomics and can be applied across a range of use cases, ensuring accurate and sensitive microbiome analysis.
Microbiome therapeutics
To qualify for clinical trials, a microbiome profiling pipeline must be demonstrably consistent in its performance. Champ was built using Clinical Microbiomics’ extensive database of more than 400,000 Metagenome Assembled Genomes (MAGs) collected across nine body sites, including but not limited to the gut, the vagina, the skin and the mouth.
Champ leverages the company’s proprietary algorithm calibrated across thousands of analyzed samples to ensure results are robust to the natural diversity across human microbiome samples.
“Microbiome profiling – the process of characterizing the composition of a microbiome sample – is a significant challenge. While there are nearly 7,000 known human microbiome species, a typical stool sample only contains about 200-300 species,” explains Nielsen.
Champ outperformed other best-in-class microbiome profiling systems based on sensitivity and false detection signals.
"This necessitates a precise method that, at the same time, is sensitive enough to detect even very low-abundance species. To further complicate matters, each sample contains strains with unique genetic make-up.”
“Therefore, we had to implement innovative solutions that made Champ sensitive, specific and yet resilient to strain variations. I’m particularly proud that we almost eliminated false signals while maintaining best-in-class recall (sensitivity),” Nielsen says.
Meanwhile, Huazhong University of Science and Technology researchers unveiled Meta-Sorter, an AI-based program that leverages neural networks to improve biome labeling for thousands of microbiome samples.
Microbiome profiling features
The features that define Champ are its high sensitivity to detect low abundant and rare species in samples and superior specificity to detect species present in samples. The program also provides correct abundance estimations and accurate taxonomic annotation of microorganisms at species resolution.
According to the company, developing microbiome therapeutics requires best-in-class microbiome profiling. Still, the immense potential of microbiome research has yet to be held back by the challenges of translating shotgun sequencing data into actionable insights.
False signals are detrimental in developing new probiotics and microbiome-based therapeutics, with incorrect bacterial candidates brought forward into costly clinical trials and their mechanisms of action misunderstood.
Champ has outperformed current microbiome profiling pipelines in rigorous benchmarking experiments.
Compared to MetaPhlAn4, Champ exhibited a 16% increase in sensitivity across various human body sites. It demonstrated a 400-fold lower false detection signal than MetaPhlAn4, Centrifuge, Kraken and Bracken.
By Inga de Jong
