# Ground truth data set of Gas Chromatography Mass Spectrometry (GCMS) analysed synthesised methylamphetamine

## TL;DR

This paper presents a dataset of GCMS analyses from 152 synthesized methylamphetamine samples using five different methods, useful for characterizing impurities and training machine learning models.

## Contribution

The novel contribution is a comprehensive GCMS dataset of methylamphetamine impurities from five synthesis routes, enabling automated detection and method characterization.

## Key findings

- GCMS data from 152 methylamphetamine samples synthesized via five methods were collected.
- The dataset supports automated detection of synthesis methods using machine learning.
- The data is structured using a standardized template for accessibility and automation.

## Abstract

Controlled substances are typically subjected to chemical analysis for the purpose of identifying them and, in certain instances, determining their purity using a comprehensive characterisation. This requires the analysis of chemical impurities that may be present in a drug sample due to the synthesis process. This process, known as impurity or drug profiling, can be applied to drugs that are chemically synthesised and/or subsequently chemically modified. The profiling of impurities can offer insights into the synthetic methods employed and, at times, the initial chemicals utilised. Our article focuses on data obtained from the repetitive synthesis (n = 152) of methylamphetamine using five distinct synthetic routes or pathways: Rosenmund Reduction, Birch Reduction, Moscow route, Hypophosphorous route and Emde route. Each sample has been analysed using Gas Chromatography Mass Spectrometry (GCMS) producing a chromatogram, which is presented in various formats.

A template for publishing GCMS data is given in [1] which provides a systematic and robust method to organise GCMS data that is amenable for automated handling by data scientists and accessible to practitioners. The template has been previously applied to MDMA [2]. This paper is the third in the series and is addressing the impurity of methylamphetamine samples using the five pathways listed above. The data can be used for methylamphetamine characterisation according to the synthesis method used, the automated detection of the synthetic method using machine learning, and for training purposes.