Cell fishing: A similarity based approach and machine learning strategy for multiple cell lines-compound sensitivity prediction

Abstract

The prediction of cell-lines sensitivity to a given set of compounds is a very important factor in the optimization of in-vitro assays. To date, the most common prediction strategies are based upon machine learning or other quantitative structure-activity relationships (QSAR) based approaches. In the present research, we propose and discuss a straightforward strategy not based on any learning modelling but exclusively relying upon the chemical similarity of a query compound to reference compounds with annotated activity against cell lines. We also compare the performance of the proposed method to machine learning predictions on the same problem. A curated database of compounds-cell lines associations derived from ChemBL version 22 was created for algorithm construction and cross-validation. Validation was done using 10-fold cross-validation and testing the models on new data obtained from ChemBL version 25. In terms of accuracy, both methods perform similarly with values around 0.65 across 750 cell lines in 10-fold cross-validation experiments. By combining both methods it is possible to achieve 66% of correct classification rate in more than 26000 newly reported interactions comprising 11000 new compounds. A Web Service implementing the described approaches (both similarity and machine learning based models) is freely available at: http://bioquimio.udla.edu.ec/cellfishing.

Bioactivity Comparison across Multiple Machine Learning Algorithms Using over 5000 Datasets for Drug Discovery

“…41 ChEMBL version 22 was used to extract molecules tested in 758 cell lines which was then used for machine learning with SVM and similarity analysis which illustrated comparable data on 10-fold cross-validation (accuracy = 0.65), but SVM was superior on external validation. 42 ChEMBL version 22 was used for compound−target interactions, and 1720 targets were selected with at least 10 compounds. Various molecular fingerprints and machine learning (Nai ̈ve Bayes and DNN) were compared with the nearest neighbor approaches.…”

Females and Males Show Differences in Early-Stage Transcriptomic Biomarkers of Lung Adenocarcinoma and Lung Squamous Cell Carcinoma

“…In summary, the dataset was randomly split into 10 equally sized sub-datasets. On each cross-validation iteration, nine sub-datasets were used to train a model and the remaining one sub-dataset was employed as the test set [ 42 ]. The overall prediction result was calculated through the results of 10 iterations.…”

A Very Large-Scale Bioactivity Comparison of Deep Learning and Multiple Machine Learning Algorithms for Drug Discovery

“…As a database like ChEMBL is so large it becomes a resource for many types of machine learning projects [37][38][39][40][41][42][43][44][45][46] . It contains a large number of cell lines with cytotoxicity data and cancer related targets and these have been combined and used recently to derive combined perturbation theory machine learning models 47 , however the authors did not explore prospective external testing of this method.…”

“…Multitask learning outperformed single task learning when the targets are similar (ROC > 0.8) but when this is not considered or they are diverse then single task learning was superior (ROC 0.76-0.79)39 . ChEMBL version 22 was used to extract molecules tested in 758 cell lines which was then used for machine learning with SVM and similarity analysis which illustrated comparable data on 10-fold cross-validation (accuracy = 0.65) but SVM was superior on external validation40 . ChEMBL version 22 was used for compound-target interactions and 1720 targets were selected with at least 10 compounds.Various molecular fingerprints and machine learning (naïve Bayes and DNN) were compared with nearest neighbor approaches.…”