Novel Algorithms for Comprehensive Untargeted Detection of Doping Agents in Biological Samples.

The researchers developed a new approach to detect the use of drugs for performance enhancement in sports. They tested fifty-seven drugs by spiking them into blank horse blood plasma and then used liquid chromatography and high-resolution mass spectrometry to analyze the composition. The majority of the drugs were accurately detected by this method, which is the first time such a comprehensive undirected drug detection method has been developed. As such, it represents a significant advance for artificial intelligence applications in sports doping control.

Objective of the study

• The aim was to address the difficulties of targeted analytical methods that are only capable of detecting known doping substances. The researchers sought to create a methodology that permits unreliable drug detection to bolster anti-doping measures in sports.

Methods used in the study

• Fifty-seven drugs were added to blank equine plasma and treated as unknowns, disregarding their exact mass and chromatographic retention times.
• The drugs in the plasma samples were then extracted and analyzed using a scientific process called liquid chromatography, coupled with high-resolution mass spectrometry (LC-HRMS).
• The raw data generated from the LC-HRMS was dealt with using metabolomics software that permitted compound detection and identification.

Results of the study

• A mathematical model was built on the resultant data and two algorithms were constructed to calculate the ratio of the mean (ROM) and the outlier index (OLI).
• The majority of the spiked drugs were accurately detected by name (52 of 57) or by chemical formula (1 of 57), showcasing the precision of the methodology.
• The detection limit for the drugs ranged from the tens of picograms to nanograms per milliliter, demonstrating the sensitivity of the new method.
• Xenobiotics and endogenous substances, substances that are made inside the body, relevant to doping control were also successfully identified using this untargeted method after they had been extracted from real-world sports samples. This validated the integrity and utility of the unreliable drug detection (UDD) methodology.

Significance of the study

• This research represents a significant advance towards using artificial intelligence for the detection of both known and emerging doping agents in sports.
• This is the first entirely unreliable drug detection (UDD) methodological approach, indicating a pioneering development in the field.