Good day, Jitendra Kuldeep. I'm Arjun. I've had a look through your background and your professional record. I'm interested in hearing more about your work and how you see yourself fitting in here with our faculty. Ready to start?

Good morning. My name is Jitendra Kuldeep and I hold a PSD in, uh, bioinformatics from CSI Central Drug Research Institute, Lucknow, uh, where my work, uh, focuses, uh, focused on uh, uh, including Ai-based methods in computational and drug discovery projects. And also my research experience expands from national and international experience in applying AI based.

In computational and cancer related datasets and currently I am working as a project scientist at ICGB where again my work focuses on application of AIML based approaches in enzyme engineering.

Thank you for sharing that, Professor Kuldeep.

Thank you for your patience, Professor Kuldeep. Let us begin by exploring your qualifications and expertise as they align with the role of a Cancer Bioinformatics Professor. Can you describe your research expertise in cancer bioinformatics and the specific areas where you have made significant contributions?

Yeah, in specifically in cancer biointhematics, actually one of the postdoc which I have done in France. So it focused on retrieving data for cancer data from NCI 60. So it consists of 60 cell lines data of cancer patients and for each of the cancer cells there were around 50,000. Molecules. So it was activity value for each of the cell line.

For a given compound and I utilized AI based advanced algorithms to make a basic classification model and also deep learning model. So utilizing those models we virtually screen large database to find the candidate molecules against those cancer cell lines.

That is quite intriguing, Professor Kuldeep. It appears you’ve adeptly combined cancer bioinformatics with AI-driven methodologies. To probe deeper, can you share how you validated the predictive models you developed during this research? Specifically, what metrics or approaches did you use to ensure their reliability?

Yeah, for the validation we use several cross validation methodologies. One of them we used was leave 1 dissimilar out methodology. So this is in short it is called as LDMO leave dissimilar molecules out. So in that all of the molecules we have done the clustering, OK and clustering actually. As a as a result of clustering, we got around 8 clusters and had each time of the training. One of the clusters was used as a test set.

And the metrics we used was PCC that is PSN correlation coefficient. Also we checked for the RMC that is root mean squared error and and sometimes R square as well. So and based on that whatever the activity value were predicted, we again applied a classification based approach by setting a clear.

Clear. By setting a clear activity threshold and based on that we also calculated the MCC, that is Matthew's correlation coefficient and other similar metrics.

Thank you.

Thank you for detailing your validation approach, Professor Kuldeep. Your thorough application of clustering, cross-validation techniques, and diverse metrics like PCC, RMSE, R-squared, and MCC underlines your rigorous methodology. Now, shifting to teaching—how would you approach teaching a complex topic like AI in cancer bioinformatics to a diverse group of graduate students, ensuring clarity and engagement?

The not only the theoretical aspect but also they will apply their knowledge in practical aspect. So they will build an intuition of the biological problem and also before that I will motivate them why we are using the AI based method in our projects. So in total I can say that a project based learning approach OK and at different grades of categorization would definitely help diverse group of students to learn the concepts in a better way.

Yeah, as you mentioned diverse group of students, so I would assume that some of them would have a low level skill, some of them might have little bit high level skills. So what I will do is like I will categorize and those means the projects. I can especially say that many projects and categorize it into basic, medium and advanced. OK. So in that way a student will be learning.

Thank you, Professor Kuldeep. Your project-based learning strategy, tailored to varying skill levels, reflects a thoughtful and inclusive teaching philosophy. Prioritizing both theoretical understanding and practical application certainly fosters deeper engagement. To extend this further, how would you evaluate your students' grasp of the concepts, particularly in a subject as interdisciplinary as this? Could you outline any specific evaluation methods or tools you would employ?

Any of the important concepts then it will be again iterated back means it will be iterative learning process. Whatever the theoretical part will be taught based and also followed by the practical approach, then followed by the assessment if they are missing then again back to the iterative cycle of theory practical and again quizzing and test. So I think this will enable the student to learn the concept in a very rigorous and very fast and very efficient way.

Understood, Professor Kuldeep. Your structured iterative method that combines theoretical instruction, practical application, and continuous assessment exhibits a well-rounded approach. Incorporating quizzes, mentor-guided projects, and iterative reinforcement ensures conceptual clarity for all levels of learners. Shifting gear slightly, can you elaborate on your experience with guiding research projects and how you foster independent inquiry in your students?

Uh, actually, uh, during my PhD days, I was involved in, uh, skill development program where I used to, uh, guide students and teach, uh, the external participants coming to enhance their skills. So in that manner, I have actually guided them and teach them, taught them all the basic concepts of AI applied to the bioinformatics and computation biology problems. Also to my PhD juniors, I have mentored them. And I have guided them in their and helped them in their projects completion. So based on this experience, I will use these kind of experience in my classroom settings to the students of diverse background. So this will definitely help them to build intuition and motivation and structured. Guidance from me so that they can learn in a very efficient way for all the concepts.

Thank you, Professor Kuldeep. It’s clear that you bring valuable experience from mentoring and skill-development initiatives, particularly in applying AI to bioinformatics challenges. Your emphasis on structured guidance and fostering intuition strongly aligns with the mentorship aspect of this role. Can we now discuss your publication history? Could you summarize key contributions or findings from your research that were published in reputed journals?

Yes. So in total I have published 16 research papers till now and some of them are in in queue in communication and some are in under preparation. But I will talk about my published work. So recently like I have published one of the paper where I have used the de Novo method of basically called as generative AI. So generative artificial intelligence has been used in. Drug target for which very low inhibited datasets was available. So we use this generative AI method to generate more number of compounds into the chemical space. And then at the following all the standard pipelines of of AI driven drug discovery, I actually discovered 1 molecule at the end which was containing different scaffolds than the. Available compounds so.

I believe diversity is very important in case of drug discovery because many of the times we see drug registers problem arising in many of the organisms. So if due to that we focused on this problem, another problem we have focused where we have used again basic classification based methodology in infectious diseases and one of them I have used in specifically if I speak.

I have used in my covectomy tuberculosis. So the drug target I have used. So I have retrieved the data set from the respective database and then classification based approach I have built and the algorithm was used and then virtual screening was performed on the database containing small molecule potential lead like molecules and then at the end of the pipeline.

With experimental validation was also used and in that we obtained out of 10 validated. Out of 10 experimental validation, we obtained 2 to 3 compounds showing very high inhibition rate in that particular Organism. So in this way there are other studies also for which we have done experimental validation in various publications.

Thank you, Professor Kuldeep. Your publication record is impressive, particularly your innovative application of generative AI to drug discovery and the emphasis on chemical diversity to address drug resistance. It is also commendable that your research involved experimental validation, which bridges computational work with tangible outcomes. Could you now elaborate on your experience with industry projects or consultancy, if any, and how that exposure might inform your teaching or research activities?

Actually I have not been directly involved in means directly means not under industry, but I have been in collaboration with industry project with a company. So it was a pharmaceutical company. So they had a project and the target was human turd. So they have the target of identifying. Potential compound which can inhibit that particular target. So in that project we have actually. We, uh, used computational methods to, uh, identify potential compounds, which actually helped, uh, a lot. And then in the wet lab validation, they found those compounds to be active. So the initially the funding, uh, was going to stop that year, but since they found some active compounds, so they extended the project to another year. So this accomplishment, I would, uh, means I would like to strengthen my profile that yes, we have.

To extended a project which was about to end and because of our computational method, they have found the promising candidate molecule and have extended the project and, uh, so, and regarding the in, uh, teaching aspect, which you asked, uh, so that I already mentioned that I have, uh, mentored my PhD juniors and external participants coming for their skill development program and, uh, I have received very good feedback from them like, but.

Uh, in that sense, uh, this teaching and, uh, research accomplishments, uh, reflects my strength in my profile.

Thank you, Professor Kuldeep. Your collaboration with the pharmaceutical industry, leading to tangible outcomes and even an extension of project funding, highlights the impactful integration of computational methods with applied research. It is commendable that you connect these achievements with your mentoring and teaching efforts, indicating a commitment to translating high-level research into effective guidance for your students. Lastly, before we conclude this section, could you speak to your future research aspirations or projects you wish to pursue, especially in alignment with cancer bioinformatics or broader interdisciplinary initiatives?

Yeah, sure. So like my career goal is and aspiration first of all is to become a committed faculty member and also involving the interdisciplinary research and also student mentoring and guided projects to the students and developing their skills and all. But and I have to make a balance between these two research and teaching. So my research objective would be like. Using AI based approaches for current scenarios, problems, whatever is there. But more specifically like you mentioned cancer by antibodies. So the target will be drug target will be from cancer domain and I will be very I'm keen to use my research experience whatever I have gained for my PhD and national and international. Research experience. So those AI based experiences will be.

Definitely useful for me in applying to the cancer related projects and finding at last a novel compound inhibitor from which can be diverse from the existing.

Existing available known therapeutics and of course I will look for, for this I will look for an collaboration, interdisciplinary research collaboration within and vit and also.

Uh, explore potential research collaboration across the globe with my existing, uh, existing potential, uh, professional networks. So, and, uh, also I will take suggestions from my senior colleagues in executing the projects and, uh, this will definitely help me in achieving my target.

Thank you, Professor Kuldeep. Your research aspirations, specifically focusing on AI-driven discovery of novel and diverse cancer therapeutics, are very compelling. Your emphasis on fostering interdisciplinary and global collaborations, along with leveraging your prior experience, demonstrates a well-rounded vision for impactful research and teaching contributions. This concludes the focus on technical competencies. Let us now move to verifying your skills and qualifications systematically. Let’s start by verifying your proficiency in Cancer Bioinformatics. Particularly, how do you approach integrating omics data—such as genomic, transcriptomic, or proteomic data—in cancer research to derive meaningful insights?

Yeah. So the various omics data are available for cancer, cancer biology. So harnessing these these data for my study will like for example single cell or from bulk RNS data at last whatever running the bandwidth pipeline will give us some marker genes and probable drug targets for those particular. Those those particular class of cancer.

So probably those can be a noble drug target not explored yet before. So exploring all the omics data will will particularly give us.

At last, drug target unexplored probably and exploring those targets using structure by informatics approach and performing virtual screening using large database for example, enamine could probably give us a novel compounds against that.

Truck target. So combining both omics data as well as structure by informatics, which is the physics based approach would definitely, uh, a very robust way of achieving uh, candidate molecule against a given drug target.

Thank you, Professor Kuldeep. Your approach of integrating omics data, identifying novel drug targets, and combining them with structure-based bioinformatics is indeed robust and well-aligned with advancing cancer therapeutics. Let us next examine your experience in teaching theory and laboratory courses. Can you walk me through a typical laboratory session you might design for graduate students studying cancer bioinformatics? Specifically, what techniques or tools would you emphasize?

Yeah, yeah. So for example, designing mini projects would definitely means specifically staying project based learning in a laboratory session would definitely help students to learn the practical aspects of the problem. So tools and techniques. For example, if we are talking about the structure based approach, then how to retrieve the data from a particular database for example. Structures are available at PDB and sequence is available from unipod if the structure is not there.

Then we can use the, uh, binary tools, for example, modeler. Now alpha fold is used to model the protein structure. And then further, we can use other techniques like uh, for example, docking tools, like freely available tools. Are there auto doc vena? Is there even Ismina is there and and how to dock the particular proper compounds into the active site of the drug target? And how does? How it? Should be interpreted, interpreted to to result interpretation is the most important thing. Otherwise the pipeline would be meaningless. So running the full pipeline along with result interpretation and then it will build the intuition to to the students and of course they will understand in this way the biological problem we are actually focusing on and how bioinformatics approaches are helpful in the earlier stages of the drug discovery project. So tools and techniques, for example.

Like I mentioned, the modeler who can be there and also Autodock Vina can be there. And if molecular dynamic simulation grow Max can be employed for that and another project, for example, if we are focusing on omics data, for example, single cell, then freely available packages are available in RCART packages there. If students, we can make them aware of Python because nowadays in the in the industry settings, Python is more.

Uh, preferred then I, I and then are so in Python package scan π is there. So using these packages, uh, students, uh, will be made, uh, uh, they will be made, uh, available of the these packages, how they are used in preprocessing and how they are used in overall processing of the data. The whole process goes from like quality control. That is how it is. The Sirat object is there.

How is scanned pies? Yeah, actually doing the internals of their how the cell clustering is being done, how the cell annotation is being done in the whole pipeline, how the DG analysis is being done, That is a differential gene expression analysis in single cell. We used basically mask can be used and also sometimes Wilcox and T test can be also used. So these are some of the techniques which can actually be used in the project and the students can explore these particular tools and techniques to enhance their knowledge.

Build intuition of the biological problem.

Yes, if a student who has learned the project, if if we will ask him to explain the project. So basic thing is if he's able to explain the project in a very simple and understandable way, which in the class if every other student is also able to grasp those things, it means the students have actually understood and understood and have gained a very good expertise of that particular project. So another thing is assessment via quizzes and tests can be done and regular. Exams.

Thank you, Professor Kuldeep. Your detailed plan for laboratory sessions demonstrates a strong commitment to equipping students with both theoretical understanding and hands-on expertise. Your inclusion of widely used tools like PDB, AlphaFold, Autodock Vina, and Python-based packages, along with advanced techniques like molecular dynamics and single-cell omics analysis, reflects excellent preparation for students entering both academia and industry. Now, moving on to student evaluation, could you share an example of how you assess student understanding during coursework or practical sessions to ensure that key learning objectives are achieved?

And regular diverse projects would be can be introduced for a single problem so that they can broaden their understanding of that particular domain. So this can be incorporated in the pipeline so that it can be efficient and useful for the students.

Thank you, Professor Kuldeep. Your emphasis on gauging students' ability to simplify and communicate their learnings, paired with assessments like quizzes, exams, and diverse project-based evaluations, demonstrates a well-rounded and student-focused methodology. Next, let us address guiding student projects and research. Could you describe how you mentor students to independently pursue research, while ensuring they remain aligned with project goals and objectives?

Yeah, setting clear timelines and goals and also making them aware of to decide what priorities they should focus on. So this will definitely help them in achieving their goal and understanding the project. So the timeline based structured learning can be introduced so that they can train, they can be trained on this particular aspect and can achieve the goal.

Thank you, Professor Kuldeep. Your focus on establishing clear timelines and priorities for students reflects a structured approach to mentoring, fostering both independence and alignment with research goals. Let us now assess your communication skills in teaching. Could you provide an example of how you simplify a concept, such as protein docking, for students who might be new to the subject?

Yeah, protein docking can be explained is like. If docking is like a clock and key, so like a key fits into the pocket of the lock, similarly A ligand can be fit into the active site of the protein. So in this way we can teach the students of the concepts of docking and also if we go into the technical aspect we can tell them of the scoring functions, what is course how the function is actually.

Our

Used and how internally the ligands are being sampled and then it is actually scored. So, uh, so in that sense, so the students can be aware of the technical aspect as well as also the imagination point of view also. So students will be able to imagine based on the example and based on the technical thing, they will be able to correlate all the.

Uh, all the steps. So, uh, this can be your utilized.

To the students.

Thank you, Professor Kuldeep. Your analogy of a lock and key is an excellent way to convey the core concept of protein docking to beginners. By pairing this with technical details like scoring functions and ligand sampling, you ensure that students grasp both the conceptual and practical aspects. Now, let us assess your qualifications further. You hold a Ph.D. in Bioinformatics, which is highly relevant. Could you briefly highlight how this academic foundation helps you contribute to advanced teaching and research in cancer bioinformatics?

Yes, since I hold a PhD in bioinformatics and I have a huge research experience which is which can be demonstrated by my research contribution in peer reviewed research articles and reported international journals. So that particular knowledge can be actually translated in the teaching environment through project based learning and also you engaging students into the research activities and also so this will also. Make students learn the.

Updated, uh, things going on, uh, these days. So, uh, and how the, actually internally, how the algorithms are working. So this can be achieved when they will have a research aptitude. So research aptitude can be, uh, introduced via teaching to them so that they build an intuition not only superficially, but from algorithmic level of depth.

Thank you, Professor Kuldeep. Your strong foundation in bioinformatics, combined with your extensive publication record, positions you to bring current, research-driven insights into the classroom. Your focus on fostering research aptitude and encouraging algorithmic understanding among students highlights a forward-thinking and enriching approach to education. Next, let us discuss your record of publishing in reputed journals. Could you explain how your experience with the publication process equips you to guide students in producing high-quality research output?

Yeah. So students will be mentor guided and as I mentioned that each and every step they will be given proper guidance and wherever they find difficult to understand any concept, that concept will be explained again and again till they find it comfortable to explain it to others. So in this way engaging students into the research activities will also build their profile in the research field and also they will be industry ready.

Because the teaching the, for example, we talked about omics. So in omics, uh, and anyone notice pipelines, we talk about making a reproducible computational workflow. So these things can only be done using.

Uh, impactful question. So the, and solving those questions using biology, using bioinformatics approaches, particularly AI based methods, which we recently see that, uh, uh, those are used to solve, uh, high-level biological problem and leads to a very high publishing in a very high impact journal. For example, recently one paper was published called as niche former. So it was actually exploring the spatial transcriptomics data and they trained. A model and it was able to predict, uh, the neighboring sales of a particular given cell. It would predict the neighboring cells. So this was achieved using deep learning, deep learning method. So, and it was published in Nature Biotechnology. So, uh, so wherever there is a difficult question solving aptitude and difficult problem-solving task, it will definitely lead to high impact. Publishing capability.

Using reproducible, uh, if they are, if these things only matter if it is reproducible. So once we pick a concept, once we pick a biological question of.