A Survey of Multi‐task Learning Methods in Chemoinformatics

Sosnin, Sergey; Vashurina, Mariia; Withnall, Michael; Karpov, Pavel; Fedorov, Maxim V.; Tetko, Igor V.

doi:10.1002/minf.201800108

Cited by 65 publications

(42 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ligands observed to be active or inactive for these tasks were likely to share similar activity with the others. This allows the network to much more effectively pick up on common structural features and learn them as reported in other studies [62,63]. However, in the case where missing data is imputed as inactive, these correlations become more difficult to learn, as negative counterexamples examples are artificially introduced.…”

Section: Datasets Classificationmentioning

confidence: 98%

Building attention and edge message passing neural networks for bioactivity and physical–chemical property prediction

et al. 2020

View full text Add to dashboard Cite

Neural Message Passing for graphs is a promising and relatively recent approach for applying Machine Learning to networked data. As molecules can be described intrinsically as a molecular graph, it makes sense to apply these techniques to improve molecular property prediction in the field of cheminformatics. We introduce Attention and Edge Memory schemes to the existing message passing neural network framework, and benchmark our approaches against eight different physical-chemical and bioactivity datasets from the literature. We remove the need to introduce a priori knowledge of the task and chemical descriptor calculation by using only fundamental graph-derived properties. Our results consistently perform on-par with other state-of-the-art machine learning approaches, and set a new standard on sparse multi-task virtual screening targets. We also investigate model performance as a function of dataset preprocessing, and make some suggestions regarding hyperparameter selection.

show abstract

Section: Datasets Classificationmentioning

confidence: 98%

Building attention and edge message passing neural networks for bioactivity and physical–chemical property prediction

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Transfer and multitask learning (TL and MTL) are two technologies that can make full use of the commonality of multiple tasks to build a more robust model . Combined with DL, they display a promising prospect in the development of novel ML‐based SFs.…”

Section: Deep Learning In Scoring Functionsmentioning

confidence: 99%

From machine learning to deep learning: Advances in scoring functions for protein–ligand docking

Shen

Ding

Wang

et al. 2019

WIREs Comput Mol Sci

164

161

View full text Add to dashboard Cite

Molecule docking has been regarded as a routine tool for drug discovery, but its accuracy highly depends on the reliability of scoring functions (SFs). With the rapid development of machine learning (ML) techniques, ML-based SFs have gradually emerged as a promising alternative for protein-ligand binding affinity prediction and virtual screening, and most of them have shown significantly better performance than a wide range of classical SFs. Emergence of more data-hungry deep learning (DL) approaches in recent years further fascinates the exploitation of more accurate SFs. Here, we summarize the progress of traditional ML-based SFs in the last few years and provide insights into recently developed DL-based SFs. We believe that the continuous improvement in ML-based SFs can surely guide the early-stage drug design and accelerate the discovery of new drugs. This article is categorized under:Computer and Information Science > Chemoinformaticsdeep learning, machine learning, molecular docking, scoring function, structure-based drug design | INTRODUCTIONTraditional drug discovery largely relies on the application of high-throughput screening, an experimental technique with acceptable performance but high cost and low efficiency. 1 With the rapid development of computational chemistry and computer technology, computer-aided drug design (CADD) has gradually emerged as a powerful technique in the design and development of new drug candidates in the past three decades. 2 Virtual screening (VS), an important branch of CADD, can enrich potential actives from large virtual compound libraries through in silico methods rather than real experiments, which can not only accelerate the process of drug discovery but also greatly reduce the time and resource cost. 3-5 Depending on whether the three-dimensional (3D) structure of a target is used or not, VS approaches can be classified into two major categories: ligand-based virtual screening (LBVS) and structure-based virtual screening (SBVS). 6 LBVS aims to discover active molecules through the models developed based on a set of known ligands of a target of interest, which may limit its capability to find novel chemotypes. Compared with LBVS, SBVS is considered to be a better choice to discover novel active compounds if the 3D structure of a given target is available. 7 Chao Shen and Junjie Ding are equivalent first authors.

show abstract

“…In this study we tried to collect and compare the QSAR models of membrane transporters to make a comprehensive transporter profiles for purposes of virtual screening of compounds for their bioavailabilty. Now point of future investigation should be to switch in new level beyond individual approaches and integrate various QSAR analyses with ontology and omics data [ 77 , 81 , 82 , 83 ]. Such simultaneously predicting with joint data analyses are shown to increase the accuracy of models by exploiting their common representation and identifying common features between individual properties, which are frequently strongly correlated with one another [ 81 , 83 ].…”

Section: Discussionmentioning

confidence: 99%

“…Now point of future investigation should be to switch in new level beyond individual approaches and integrate various QSAR analyses with ontology and omics data [ 77 , 81 , 82 , 83 ]. Such simultaneously predicting with joint data analyses are shown to increase the accuracy of models by exploiting their common representation and identifying common features between individual properties, which are frequently strongly correlated with one another [ 81 , 83 ]. For sure not in so far future a chemical characterization will be even more empowered with an artificial intelligence assisted software, which would make decisions based on integration of numerous data from in silico, in vitro, in vivo and in situ analyses [ 84 ].…”

Section: Discussionmentioning

confidence: 99%

An In Silico Approach for Assessment of the Membrane Transporter Activities of Phenols: A Case Study Based on Computational Models of Transport Activity for the Transporter Bilitranslocase

Venko

Novič

2019

Molecules

View full text Add to dashboard Cite

Phenols are the most abundant naturally accessible antioxidants present in a human normal diet. Since numerous beneficial applications of phenols as preventive agents in various diseases were revealed, the evaluation of phenols bioavailability is of high interest of researchers, consumers and drug manufacturers. The hydrophilic nature of phenols makes a cell membrane penetration difficult, which imply an alternative way of uptake via membrane transporters. However, the structural and functional data of membrane transporters are limited, thus the in silico modelling is really challenging and urgent tool in elucidation of transporter ligands. Focus of this research was a particular transporter bilitranslocase (BTL). BTL has a broad tissue expression (vascular endothelium, absorptive and excretory epithelia) and can transport wide variety of poly-aromatic compounds. With available BTL data (pKi [mmol/L] for 120 organic compounds) a robust and reliable QSAR models for BTL transport activity were developed and extrapolated on 300 phenolic compounds. For all compounds the transporter profiles were assessed and results show that dietary phenols and some drug candidates are likely to interact with BTL. Moreover, synopsis of predictions from BTL models and hits/predictions of 20 transporters from Metrabase and Chembench platforms were revealed. With such joint transporter analyses a new insights for elucidation of BTL functional role were acquired. Regarding limitation of models for virtual profiling of transporter interactions the computational approach reported in this study could be applied for further development of reliable in silico models for any transporter, if in vitro experimental data are available.

show abstract

A Survey of Multi‐task Learning Methods in Chemoinformatics

Cited by 65 publications

References 34 publications

Building attention and edge message passing neural networks for bioactivity and physical–chemical property prediction

Building attention and edge message passing neural networks for bioactivity and physical–chemical property prediction

From machine learning to deep learning: Advances in scoring functions for protein–ligand docking

An In Silico Approach for Assessment of the Membrane Transporter Activities of Phenols: A Case Study Based on Computational Models of Transport Activity for the Transporter Bilitranslocase

Contact Info

Product

Resources

About