One of the main authors, Jiangjiang Zhu, professor of human sciences at the Ohio State University, specifies that “although additional analyzes are necessary, the” biomarker discovery pipeline “is already very promising as a non -invasive diagnosis of colorectal cancer but also as a tool for monitoring the progression of the disease. Finally, the analysis of metabolic biomarkers could also be used to assess the effectiveness of treatment ”.
The study Analysis of blood samples from more than 1,000 participants and reveals:
- of the Metabolic modifications associated with the evolution of the severity of the disease and Genetic mutations Known to increase the risk of colorectal cancer;
In practice, the new platform makes it possible to extract 2 sets of biological data from the 626 blood samples analyzed, coming from participants with colorectal cancer, some of which with high -risk genetic mutations and 402 other samples from healthy witnesses, paired by age and sex:
- Metabolites, biochemical reactions that break down food to produce energy and ensure other essential functions,
- Transcripts, DNA Instructions RNA readings which reflect the associated protein modifications.
Metabolic routes linked to a family of compounds called purines, necessary for the training and degradation of DNA, are distinguished during this analysis: they are generally more active in cancer patients than in healthy witnesses, and less active at more advanced tumor stages.
If the tool is not intended to replace colonoscopy as a reference for cancer screening, if additional validation studies on larger samples remain necessary, it could be transposed in clinical environment as first -line detection test.
These works also illustrate the advancement of artificial intelligence techniques and automatic learning in health and diagnosis in particular, here with the combination of 2 technologies, discriminating analysis by the smallest partial squares which allows the overall differentiation of molecular profiles, and an artificial neural network which allows the identification of predictive molecules of the tumor.
It is therefore the promise of a new bioinformatic pipeline for the diagnosis and monitoring of colorectal cancer, with the prospect of an always better personalized management.
