TMLRpred: A machine learning classification model to distinguish reversible EGFR double mutant inhibitors.

The EGFR is a clinically important therapeutic drug target in lung cancer. The first-generation tyrosine kinase inhibitors used in clinics are effective against L858R-mutated EGFR. However, relapse of the disease due to the presence of resistant mutation (T790M) makes these inhibitors ineffective. This has necessitated the need to identify new potent EGFR inhibitors against the resistant double mutants. Therefore, various machine learning techniques ((instance-based learner (IBK), naïve Bayesian (NB), sequential minimal optimization (SMO), and random forest (RF)) were employed to develop twelve classification models on three different datasets (high, moderate, and weakly active inhibitors). The models were validated using fivefold cross-validation and independent validation datasets. It was observed that the random forest-based models showed best performance. Also, functional groups, PubChem fingerprints, and substructure of highly active inhibitors were compared to inactive to identify structural features which are important for activity. To promote open-source drug discovery, a tool has been developed, which incorporates the best performing models and allows users to predict the potential of chemical molecules as anti-TMLR inhibitor. It is expected that the machine learning classification models developed in this study will pave way for identifying novel inhibitors against the resistant EGFR double mutants.

Synthesis of metronidazole based thiazolidinone analogs as promising antiamoebic agents.

Metronidazole and its derivatives are widely used for the treatment of amoebiasis. However, metronidazole is considered as the standard drug but it has many side effects. The present study describes the synthesis of a series of metronidazole based thiazolidinone analogs via Knoevenagel condensation of 4-[2-(2-methyl-5-nitro-1H-imidazole-1-yl)ethoxy]benzaldehyde 1 with various thiazolidinone derivatives 2-14 to get the new scaffold (15-27) having better activity and lesser toxicity. Six compounds have shown better efficacy and lesser cytotoxicity than the standard drug metronidazole towards HM1: IMSS strain of Entamoeba histolytica. These compounds may combat the problem of drug resistance and might be effective in identifying potential alternatives for future drug discovery against EhOASS.

ADMET profiling of geographically diverse phytochemical using chemoinformatic tools.

Aim: Phytocompounds are important due to their uniqueness, however, only few reach the development phase due to their poor pharmacokinetics. Therefore, preassessing the absorption, distribution, metabolism, excretion and toxicity (ADMET) properties is essential in drug discovery. Methodology: Biologically diverse databases (Phytochemica, SerpentinaDB, SANCDB and NuBBEDB) covering the region of India, Brazil and South Africa were considered to predict the ADMET using chemoinformatic tools (Qikprop, pkCSM and DataWarrior). Results: Screening through each of pharmacokinetic criteria resulted in identification of 24 compounds that adhere to all the ADMET properties. Furthermore, assessment revealed that five have potent anticancer biological activity against cancer cell lines. Conclusion: We have established an open-access database (ADMET-BIS) to enable identification of promising molecules that follow ADMET properties and can be considered for drug development.

Structure-activity Relationship Study on Therapeutically Relevant EGFR Double Mutant Inhibitors.

BACKGROUND: EGFR is a clinically approved drug target in cancer. The first generation tyrosine kinase inhibitors targeting L858R mutated EGFR are routinely used to treat non-small cell lung cancer (NSCLC). However, the presence of a secondary mutation (T790M) tenders these inhibitors ineffective and thus results in the relapse of the disease. OBJECTIVE: New reversible inhibitors are required, which act against T790M/L858R (TMLR) double mutants and overcome resistance. METHOD: In the present study, various Fragment based QSAR (G-QSAR) models along with interaction terms have been studied for amino-pyrimidine derivatives having biological activity against TMLR mutant enzyme. RESULTS: The G-QSAR models developed using partial least squares regression via stepwise forward- backward variable selection technique showed the best results. The model showed a high correlation coefficient (r² = 0.86), cross-validation coefficient (q² = 0.81) and predicted correlation (predicted r² = 0.62), which indicated that the model is robust and predictive. Based on the model, it was revealed that at R1 position increasing saturated carbon (number of -CH atom connected with 3 single bonds i.e. SsssCHcount) and retention index (chi3) is desired for the enhancement of bioactivity. Additionally, at the R2 position, increasing lipophilic character (slogp) and at site R3, the polarizability of compound need to be increased for better inhibitory activity. We further studied the contribution of interactions among significant descriptors in enhancing the activity of the compounds. It revealed that the presence of Sum((R1-SsssCHcount, R2-slogp) and Mult(R1-chi3, R3-polarizabilityAHC) are the most significantly influencing descriptors. We further compared the variation in the most and least active compounds which established that retention of the above properties is essential for imparting significant inhibitory activity to these molecules. CONCLUSION: The study provides site specific information wherein chemical group variation influences the inhibitory potency of TMLR amino-pyrimidine inhibitors, which can be used for designing new molecules with the desired activity.

Exploring structural features of EGFR-HER2 dual inhibitors as anti-cancer agents using G-QSAR approach.

Simultaneous inhibition of EGFR and HER2 by dual-targeting inhibitors is an established anti-cancer strategy. Therefore, a recent trend in drug discovery involves understanding the features of such dual inhibitors. In this study, three different G-QSAR models were developed corresponding to individual EGFR, HER2 and the dual-model for both receptors. The dual-model provided site-specific information wherein (i) increasing electronegative character and higher index of saturated carbon at R4 position; (ii) presence of chlorine atom at R2 position; (iii) decreasing alpha modified shape index at R1 and R3 positions; and (iv) less electronegativity at R2 position; were found important for enhancing the dual activity. Also, comparison of dual-model with the EGFR/HER2 individual models revealed that it incorporates the properties of both models and, thus, represents a combination of EGFR/HER2. Further, fragment analysis revealed that R2 and R4 are important for imparting high potency while specificity is decided by R1/R3 fragment. We also checked the predictive ability of the dual-model by determining applicability domain using William's plot. Also, analysis of active molecules showed they show favorable substitutions that agree with the constructed dual-model. Thus, we have been successful in developing a single dual-response QSAR model to get an insight into various structural features influencing EGFR/HER2 activity.

QSAR of clinically important EGFR mutant L858R/T790M pyridinylimidazole inhibitors.

EGFR is a well-established therapeutic target of clinical relevance in cancer. However, acquisition of secondary mutation (T790M) makes first-generation inhibitors ineffective. Therefore, to circumvent the problem of resistance, new T790M/L858R (TMLR) double mutant inhibitors are required. In this study, fragment-based QSAR models (GQSAR) were generated for pyridinylimidazole derivatives having biological activity against TMLR mutants. The GQSAR model developed using partial least squares regression via stepwise forward-backward variable selection technique showed best results as judged using statistical parameters (r2 , q2 , and pred_r2 ). Additionally, applicability domain of the model was verified using Williams plot, which indicated that the predicted data are reliable. The GQSAR provided site-specific clues wherein modifications related to decreasing lipophilic character and rotatable bonds and increasing SaaCHE-index are required for improving inhibitory activity. Overall, the study indicated that the presence of acrylamide at R5 is essential for covalent bond formation with Cys797 and occurrence of aromatic residue at R2 is required for occupying hydrophobic region next to Met790 gatekeeper residue. Based on this information, new derivatives were designed that show better inhibitory activity than the experimentally reported most active molecules. Thus, the model developed can be used to design new pyridinylimidazole derivatives with improved TMLR bioactivity.

Unraveling structural requirements of amino-pyrimidine T790M/L858R double mutant EGFR inhibitors: 2D and 3D QSAR study.

EGFR is an important drug target in cancer. However, the ineffectiveness of first generation inhibitors due to the occurrence of a secondary mutation (T790M) results in the relapse of the disease. Identification of reversible inhibitors against T790M/L858R double mutants (TMLR) thus is a foremost requirement. In this study, various 2 D and 3 D Quantitative Structure-Activity Relationship models were built for amino-pyrimidine compounds with their known biological activity against TMLR mutants. The model developed using multiple linear regression statistical method via stepwise forward-backward variable selection technique showed the best results in terms of internal and external predictivity. The 2D-QSAR model indicated that the presence of electronegative atom, H-bond donors, moderate slogp, count of number of N atoms separated from O (T_N_O_4), 4pathClusterCount and number of S atom connected with two single bonds (SssSE-index), is required for increasing the inhibitory potential of compounds. Also, the 3D-QSAR model suggested that electronegative group at certain positions along with the presence of bulky groups is beneficial for good inhibition activity of the compounds. Thus, the QSAR models developed in the present work can be used for predicting the TMLR bioactivity of a new series of amino-pyrimidine derivatives. To the best of the author's knowledge, this is the first study which deals with the development of 2 D and 3D-QSAR models for double mutant TMLR inhibitors.

Insight into structural requirements of antiamoebic flavonoids: 3D-QSAR and G-QSAR studies.

Plant-based flavonoids have been found to exhibit strong inhibitory capability against Entamoeba histolytica. So, various QSAR models have been developed to identify the critical features that are responsible for the potency of these molecules. 3D-QSAR analysis using k-nearest neighbour molecular field analysis via stepwise forward-backward variable selection method showed best results for both internal and external predictive ability of the model (i.e., q2  = 0.64 and pred_r2  = 0.56). Also, a group-based QSAR (G-QSAR) model was developed based on partial least squares regression combined with stepwise forward-backward variable selection method. It gave best parametric results (r2  = 0.74, q2  = 0.56 and pred_r2  = 0.54) which implied that the model is highly predictive. 3D-QSAR established that presence/absence of bulk near rings B and C is important in deciding the inhibitory potential of these molecules. Additionally, G-QSAR provided site-specific clue wherein modifications related to molecular weight, electronegativity and separation of an oxygen atom in rings A and C can result in enhanced biological activity. To the best of the author's knowledge, this is the first QSAR study of antiamoebic flavonoids, and therefore, we expect the results to be useful in the design of more potent antiamoebic inhibitors.

Pharmacokinetic profiling of anticancer phytocompounds using computational approach.

INTRODUCTION: Natural products exhibit diverse scaffolds and are considered as suitable candidates for development of leads. However, poor pharmacokinetics often acts as a hindrance during the drug discovery process. OBJECTIVE: With a view of exploring the absorption, distribution, metabolism, excretion and toxicity (ADMET) profile of plant-based anticancer compounds, open-access databases (NPACT, CancerHSP and TaxKB) were analysed to identify molecules having properties favourable for drug ability. METHODOLOGY: Our workflow involved identification of molecules capable of passing each of the ADMET barriers based on physicochemical properties of molecules, and physiological barriers and factors. RESULTS: The results revealed that out of 5086 phytomolecules, 63% were orally absorbable and 52% distributable. Also, an appreciable proportion of these compounds (45%) could be metabolised and excreted. Furthermore, 28% were found to be non-toxic for cardio toxicity and central nervous system (CNS) activity. Additionally, comparison against known anticancer drugs (reference dataset) revealed that the three libraries exhibit similar trends, thus providing additional confidence to the predictions. Overall, 28% of the molecular dataset was found to have suitable pharmacokinetic properties. We have also discussed a few natural products which exhibit favourable ADMET as well as low nano-micromolar in vitro anticancer activity. CONCLUSION: We have created an interactive database (ADMETCan), which provides access to predicted ADMET of these anticancer phytomolecules. The ease of availability of this dataset is expected to minimise failure rate of these compounds and thus is expected to be beneficial to the scientific community involved in anticancer identification and development.