5-Year Trends in QSAR and its Machine Learning Methods

Area of Science:

• Computational Chemistry
• Medicinal Chemistry
• Cheminformatics

Background:

• Quantitative Structure-Activity Relationships (QSAR) is a foundational computational chemistry technique.
• A noticeable decline in QSAR publications has been observed in recent years.

Purpose of the Study:

• To analyze current trends in QSAR research.
• To specifically examine the adoption of machine learning methods within QSAR.

Main Methods:

• Bibliometric analysis of QSAR articles from top molecular modeling and medicinal chemistry journals.
• Comparison of publication data from 2009 and 2015 to identify shifts in methodology and focus.

Main Results:

• A twofold decrease in the proportion of QSAR studies was observed between 2009 and 2015.
• Journals showed a reduced likelihood of publishing Multiple Linear Regression models, favoring Random Forest and Naïve Bayes.
• 3D-QSAR remained prevalent, with a slight decrease in molecular modeling journals but an increase in medicinal chemistry publications.

Conclusions:

• Potential reasons for the QSAR decline include stricter journal acceptance criteria, the routine nature of methods, and field maturation.
• The integration of advanced machine learning methods is expected to revitalize QSAR's role in drug design, moving it towards wider adoption.