Exploration of good and bad structural fingerprints for inhibition of indoleamine-2,3-dioxygenase enzyme in cancer immunotherapy using Monte Carlo optimization and Bayesian classification QSAR modeling

Abstract

Indoleamine-2,3-dioxygenase 1 (IDO1) is an extrahepatic, heme-containing and tryptophan-catalyzing enzyme responsible for causing blockade of T-cell proliferation and differentiation by depleting tryptophan level in cancerous cells. Therefore, inhibition of IDO1 may be a useful strategy for immunotherapy against cancer. In this study, 448 structurally diverse IDO1 inhibitors with a wide range of activity has been taken into consideration for classification QSAR analysis through Monte Carlo Optimization by using different splits as well as different combinations of SMILES-based, graph-based and hybrid descriptors. The best model from Monte Carlo optimization was interpreted to find out the good and bad structural fingerprints for IDO1 and further justified by using Bayesian classification QSAR modeling. Among the three splits in Monte Carlo optimization, the statistics of the best model was obtained from Split 3: sensitivity = 0.87, specificity = 0.91, accuracy = 0.89 and MCC = 0.78. In Bayesian classification modeling, the ROC scores for training and test set were found to be 0.91 and 0.86, respectively. The combined modeling analysis revealed that the presence of aryl hydrazyl sulphonyl moiety, furazan ring, halogen substitution, nitro group and hetero atoms in aromatic system can be very useful in designing IDO1 inhibitors. All the good and bad structural fingerprints for IDO1 were identified and are justified by correlating these fragments to the inhibition of IDO1 enzyme. These structural fingerprints will guide the researchers in this field to design better inhibitors against IDO1 enzyme for cancer immunotherapy.

Communicated by Ramaswamy H. Sarma

Keywords:

• Indoleamine-2,3-dioxygenase
• IDO 1 inhibitors
• structural fingerprint
• Bayesian classification model
• classification QSAR
• Monte Carlo Optimization

Acknowledgments

BB is grateful to the All India Council for Technical Education (AICTE), New Delhi for awarding fellowship. SAA sincerely acknowledges Council of Scientific and Industrial Research (CSIR), New Delhi for awarding the Senior Research Fellowship [FILE NO.: 09/096(0967)/2019-EMR-I, Dated: 01-04-2019]. SAA is also thankful to Jadavpur University, Kolkata for awarding Junior Research Fellowship under State Government Fellowship Scheme of Jadavpur University, Kolkata, India. TJ is thankful for the financial support from RUSA 2.0 and UPE Phase II Program of University Grants Commission (UGC), New Delhi to Jadavpur University, Kolkata, India. Authors are thankful to the authority of Department of Pharmaceutical Sciences, Dr. Harisingh Gour University, India and Department of Pharmaceutical Technology, Jadavpur University, India for providing research facilities.

Disclosure statement

No potential conflict of interest was reported by the authors.