A neural network based classification scheme for cytotoxicity predictions:Validation on 30,000 compounds

Molnár, László; Keserü, György M.; Papp, Ákos; Lőrincz, Zsolt; Ambrus, Géza; Darvas, Ferenç

doi:10.1016/j.bmcl.2005.10.079

Cited by 18 publications

(12 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several computational methods [26,42,43,48] can predict cytotoxicity, but rely on data unavailable for most substances. We propose that topological properties of drug-affected genes (e.g., degree in PPI networks) may be valuable as additional predictive variables.…”

Section: Drugs That Affect Central Genes Are More Likely To Be Toxicmentioning

confidence: 99%

Network-based characterization of drug-regulated genes, drug targets, and toxicity

Kotlyar

Fortney

Jurišica

2012

Methods

View full text Add to dashboard Cite

Proteins do not exert their effects in isolation of one another, but interact together in complex networks. In recent years, sophisticated methods have been developed to leverage protein-protein interaction (PPI) network structure to improve several stages of the drug discovery process. Network-based methods have been applied to predict drug targets, drug side effects, and new therapeutic indications. In this paper we have two aims. First, we review the past contributions of network approaches and methods to drug discovery, and discuss their limitations and possible future directions. Second, we show how past work can be generalized to gain a more complete understanding of how drugs perturb networks. Previous network-based characterizations of drug effects focused on the small number of known drug targets, i.e., direct binding partners of drugs. However, drugs affect many more genes than their targets - they can profoundly affect the cell's transcriptome. For the first time, we use networks to characterize genes that are differentially regulated by drugs. We found that drug-regulated genes differed from drug targets in terms of functional annotations, cellular localizations, and topological properties. Drug targets mainly included receptors on the plasma membrane, down-regulated genes were largely in the nucleus and were enriched for DNA binding, and genes lacking drug relationships were enriched in the extracellular region. Network topology analysis indicated several significant graph properties, including high degree and betweenness for the drug targets and drug-regulated genes, though possibly due to network biases. Topological analysis also showed that proteins of down-regulated genes appear to be frequently involved in complexes. Analyzing network distances between regulated genes, we found that genes regulated by structurally similar drugs were significantly closer than genes regulated by dissimilar drugs. Finally, network centrality of a drug's differentially regulated genes correlated significantly with drug toxicity.

show abstract

Section: Drugs That Affect Central Genes Are More Likely To Be Toxicmentioning

confidence: 99%

Network-based characterization of drug-regulated genes, drug targets, and toxicity

Kotlyar

Fortney

Jurišica

2012

Methods

View full text Add to dashboard Cite

show abstract

“…The high-throughput in vitro assay for measuring toxicity and antiproliferative effects of small molecules was implemented as described earlier [56]. Cell lines were grown in culture flasks to 90% confluences, then harvested in counted cell density and seeded into 384-well microtiter plates.…”

Section: Methodsmentioning

confidence: 99%

A Chemocentric Approach to the Identification of Cancer Targets

Flachner¹,

Lőrincz²,

Carotti

et al. 2012

PLoS ONE

View full text Add to dashboard Cite

A novel chemocentric approach to identifying cancer-relevant targets is introduced. Starting with a large chemical collection, the strategy uses the list of small molecule hits arising from a differential cytotoxicity screening on tumor HCT116 and normal MRC-5 cell lines to identify proteins associated with cancer emerging from a differential virtual target profiling of the most selective compounds detected in both cell lines. It is shown that this smart combination of differential in vitro and in silico screenings (DIVISS) is capable of detecting a list of proteins that are already well accepted cancer drug targets, while complementing it with additional proteins that, targeted selectively or in combination with others, could lead to synergistic benefits for cancer therapeutics. The complete list of 115 proteins identified as being hit uniquely by compounds showing selective antiproliferative effects for tumor cell lines is provided.

show abstract

“…In another study, in-house cytotoxity data obtained for a diverse set of 30 000 compounds was used to classify / predict compounds as cytotoxic / noncytotoxic [58]. This ANN based approach used a training set of 12 998 most and least-toxic compounds and atomic fragmental descriptors as implemented in Pallas PrologP (CompuDrug International, South San Francisco, CA, USA) program.…”

Section: Cytotoxicitymentioning

confidence: 99%

Two-Dimensional (2D) In Silico Models for Absorption, Distribution, Metabolism, Excretion and Toxicity (ADME/T) in Drug Discovery

Kharkar

2010

CTMC

View full text Add to dashboard Cite

With the dawn of new century, major technological advances in the drug discovery field have revolutionized absorption, distribution, metabolism, excretion and toxicity (ADME/T) profiling of new chemical entities (NCEs) among others. The progress made in the in vitro experimental determination of the ADME/T properties fueled the growth in the so-called predictive ADME/T. The process of in silico model development improved significantly with the availability of high quality data as well newer, more accurate statistical methods of analysis. Even several such models appear in the literature regularly, the prediction accuracy is limited to 'local' rather than 'global' applicability domain in majority of the cases. Majority of the efforts are still needed to address several issues such as data quality, accuracy of the results, etc., to increase the usefulness of these models. The ultimate aim is to develop in silico ADME/T models which will largely replace their in vitro experimental counterparts. The current review article discusses the two-dimensional (2D) approaches used in the predictive ADME/T model development and their limitations and usefulness in the discovery process.

show abstract

A neural network based classification scheme for cytotoxicity predictions:Validation on 30,000 compounds

Cited by 18 publications

References 8 publications

Network-based characterization of drug-regulated genes, drug targets, and toxicity

Network-based characterization of drug-regulated genes, drug targets, and toxicity

A Chemocentric Approach to the Identification of Cancer Targets

Two-Dimensional (2D) In Silico Models for Absorption, Distribution, Metabolism, Excretion and Toxicity (ADME/T) in Drug Discovery

Contact Info

Product

Resources

About