Small molecule retrieval from tandem mass spectrometry: what are we optimizing for?
This work addresses a key challenge in computational mass spectrometry for researchers in chemistry and biology, though it is incremental as it builds on existing deep learning methods.
The study investigates the impact of different loss functions on deep learning models for small molecule identification from mass spectrometry data, revealing a fundamental trade-off between accurate fingerprint prediction and effective molecular retrieval.
One of the central challenges in the computational analysis of liquid chromatography-tandem mass spectrometry (LC-MS/MS) data is to identify the compounds underlying the output spectra. In recent years, this problem is increasingly tackled using deep learning methods. A common strategy involves predicting a molecular fingerprint vector from an input mass spectrum, which is then used to search for matches in a chemical compound database. While various loss functions are employed in training these predictive models, their impact on model performance remains poorly understood. In this study, we investigate commonly used loss functions, deriving novel regret bounds that characterize when Bayes-optimal decisions for these objectives must diverge. Our results reveal a fundamental trade-off between the two objectives of (1) fingerprint similarity and (2) molecular retrieval. Optimizing for more accurate fingerprint predictions typically worsens retrieval results, and vice versa. Our theoretical analysis shows this trade-off depends on the similarity structure of candidate sets, providing guidance for loss function and fingerprint selection.