Genetic-Driven Druggable Target Identification and Validation

Floris, Matteo; Olla, Sergio; Schlessinger, David; Cucca, Francesco

doi:10.1016/j.tig.2018.04.004

Cited by 52 publications

(34 citation statements)

References 94 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, in order to apply an unbiased approach to score every possible protein, the method described here lacks of features based on genomic evidence [38,39] as these features could not be extended to the whole proteome. Another important limitation of our method is that it is completely indirect, i.e.…”

Section: Discussionmentioning

confidence: 99%

Machine learning prediction of oncology drug targets based on protein and network properties

Dezső¹,

Ceccarelli

2020

BMC Bioinformatics

View full text Add to dashboard Cite

Background: The selection and prioritization of drug targets is a central problem in drug discovery. Computational approaches can leverage the growing number of large-scale human genomics and proteomics data to make insilico target identification, reducing the cost and the time needed. Results: We developed a machine learning approach to score proteins to generate a druggability score of novel targets. In our model we incorporated 70 protein features which included properties derived from the sequence, features characterizing protein functions as well as network properties derived from the protein-protein interaction network. The advantage of this approach is that it is unbiased and even less studied proteins with limited information about their function can score well as most of the features are independent of the accumulated literature. We build models on a training set which consist of targets with approved drugs and a negative set of non-drug targets. The machine learning techniques help to identify the most important combination of features differentiating validated targets from non-targets. We validated our predictions on an independent set of clinical trial drug targets, achieving a high accuracy characterized by an Area Under the Curve (AUC) of 0.89. Our most predictive features included biological function of proteins, network centrality measures, protein essentiality, tissue specificity, localization and solvent accessibility. Our predictions, based on a small set of 102 validated oncology targets, recovered the majority of known drug targets and identifies a novel set of proteins as drug target candidates. Conclusions: We developed a machine learning approach to prioritize proteins according to their similarity to approved drug targets. We have shown that the method proposed is highly predictive on a validation dataset consisting of 277 targets of clinical trial drug confirming that our computational approach is an efficient and costeffective tool for drug target discovery and prioritization. Our predictions were based on oncology targets and cancer relevant biological functions, resulting in significantly higher scores for targets of oncology clinical trial drugs compared to the scores of targets of trial drugs for other indications. Our approach can be used to make indication specific drug-target prediction by combining generic druggability features with indication specific biological functions.

show abstract

Section: Discussionmentioning

confidence: 99%

Machine learning prediction of oncology drug targets based on protein and network properties

Dezső¹,

Ceccarelli

2020

BMC Bioinformatics

View full text Add to dashboard Cite

show abstract

“…Furthermore, our results suggest that the same variant associates with both disease course and immunological parameters (TNF, RANTES and IL-5). This “coincident association” between disease and intermediate phenotypes is known to increase the interest in any development linked to the therapeutic modulation or monitoring of the intermediate phenotypes themselves [ 39 , 40 ]. These considerations warrant further investigations on how genetic variants of ADA affect the enzyme’s activity (together with the above immunological parameters) in MS patients and further studies exploring different SNPs of ADA gene are required to better characterize the relationship between ADA genetic variability and MS pathogenesis.…”

Section: Discussionmentioning

confidence: 99%

A Single Nucleotide ADA Genetic Variant Is Associated to Central Inflammation and Clinical Presentation in MS: Implications for Cladribine Treatment

Bassi

Buttari

Simonelli

et al. 2020

Genes

View full text Add to dashboard Cite

In multiple sclerosis (MS), activated T and B lymphocytes and microglial cells release various proinflammatory cytokines, promoting neuroinflammation and negatively affecting the course of the disease. The immune response homeostasis is crucially regulated by the activity of the enzyme adenosine deaminase (ADA), as evidenced in patients with genetic ADA deficiency and in those treated with cladribine tablets. We investigated in a group of patients with MS the associations of a single nucleotide polymorphism (SNP) of ADA gene with disease characteristics and cerebrospinal fluid (CSF) inflammation. The SNP rs244072 of the ADA gene was determined in 561 patients with MS. Disease characteristics were assessed at the time of diagnosis; furthermore, in 258 patients, proinflammatory and anti-inflammatory molecules were measured in the CSF. We found a significant association between rs244072 and both clinical characteristics and central inflammation. In C-carriers, significantly enhanced disability and increased CSF levels of TNF, IL-5 and RANTES was observed. In addition, lower CSF levels of the anti-inflammatory cytokine IL-10 were found. Finally, the presence of the C allele was associated with a tendency of increased lymphocyte count. In MS patients, ADA SNP rs244072 is associated with CSF inflammation and disability. The selective targeting of the ADA pathway through cladribine tablet therapy could be effective in MS by acting on a pathogenically relevant biological mechanism.

show abstract

“…Finally, in order to apply an unbiased approach to score every possible protein, the method described here lacks of features based on genomic evidence [33,34] as these features could not be extended to the whole proteome, however we believe that the target druggability score developed here is a complementary approach with respect to the ones based on genetic evidence and can be easily integrated with these more focused studies.…”

Section: The Interaction Between Drugs and Their Targets Activates Simentioning

confidence: 99%

Machine learning prediction of oncology drug targets based on protein and network properties

Dezső

Ceccarelli

2019

Preprint

View full text Add to dashboard Cite

Background The selection and prioritization of drug targets is a central problem in drug discovery. Computational approaches can leverage the growing number of large-scale human genomics and proteomics data to make in-silico target identification, reducing the cost and the time needed. Results We developed a machine learning approach to score proteins to generate a druggability score of novel targets. In our model we incorporated 70 protein features which included properties derived from the sequence, features characterizing protein functions as well as network properties derived from the protein-protein interaction network. The advantage of this approach is that it is unbiased and even less studied proteins with limited information about their function can score well as most of the features are independent of the accumulated literature. We build models on a training set which consist of targets with approved drugs and a negative set of non-drug targets. The machine learning techniques help to identify the most important combination of features differentiating validated targets from non-targets. We validated our predictions on an independent set of clinical trial drug targets, achieving a high accuracy characterized by an AUC of 0.89. Our most predictive features included biological function of proteins, network centrality measures, protein essentiality, tissue specificity, localization and solvent accessibility. Our predictions, based on a small set of 102 validated oncology targets, recovered the majority of known drug targets and identifies a novel set of proteins as drug target candidates. Conclusions We developed a machine learning approach to prioritize proteins according to their similarity to approved drug targets. We have shown that the method proposed is highly predictive on a validation dataset consisting of 277 targets of clinical trial drug confirming that our computational approach is an efficient and cost-effective tool for drug target discovery and prioritization. Our predictions were based on oncology targets and cancer relevant biological functions, resulting in significantly higher scores for targets of oncology clinical trial drugs compared to the scores of targets of trial drugs for other indications. Our approach can be used to make indication specific drug-target prediction by combining generic druggability features with indication specific biological functions.

show abstract

Genetic-Driven Druggable Target Identification and Validation

Cited by 52 publications

References 94 publications

Machine learning prediction of oncology drug targets based on protein and network properties

Machine learning prediction of oncology drug targets based on protein and network properties

A Single Nucleotide ADA Genetic Variant Is Associated to Central Inflammation and Clinical Presentation in MS: Implications for Cladribine Treatment

Machine learning prediction of oncology drug targets based on protein and network properties

Contact Info

Product

Resources

About