Maharashtra State Board of Technical Education,
Mumbai
A
SEMINAR REPORT ON
“Drug Discovery Optimization Using
Reinforcement Learning”
Submitted By
Atmaja Dattatray More
Under The Guidance Of
Prof. Nawale S.K
DEPARTMENT OF COMPUTER
ENGINEERING
SAMARTH POLYTECHNIC, BELHE
TAL-JUNNAR, DIST-PUNE, PIN-
412410
(2025 – 2026)
Maharashtra State Board of Technical Education,
Mumbai
SEMINAR REPORT
“Drug Discovery Optimization Using Reinforcement Learning”
as a partial fulfilment of requirement of the
THIRD YEAR DIPLOMA IN
Computer Engineering
Submitted By
Name Of Student Enrollment Number
Atmaja Dattatray More 23212310193
For The Academic Year 2025-26
Mr. Nawale S.K Mr. Kapile A. S
(Head Of Department) (Principal)
Mr. Nawale S.K
(Internal Guide) (External Guide)
ACKNOWLEDGEMENT
First I say thanks to our respected Principal Mr. Kapile A. S. who always Inspire and
Cheer us for achieving the goals and objectives. His right guidelines and valuable
suggestions motivate me for doing my Duties and Responsibilities.
We would like to express our heartfelt gratitude towards those who where
instrumental in the successful development of our Seminar Topic.
We thanks to Mr. Nawale S. K. (Head of Department and Internal Guide) for
providing us with adequate facilities, ways and means by which was able to complete the
Seminar Report on “Drug Discovery Optimization Using Reinforcement Learning”.
Without whose valuable guidance and support this project could not have been
accomplished.
We also sincerely acknowledge the help of the entire staff of the Computer
Department and our classmates whose co-operation has made our project a success.
Atmaja Dattatray More
ABSTRACT
The seminar titled “Drug Discovery Optimization Using Reinforcement Learning” explores
the integration of Artificial Intelligence, specifically Reinforcement Learning (RL), into the
drug discovery process. Traditional drug discovery is a lengthy, expensive, and complex
procedure that often takes over a decade to bring a single drug to market. Challenges such as
high failure rates, time-intensive experimentation, and resource constraints necessitate
innovative computational approaches.
Reinforcement Learning offers a promising solution by enabling automated and intelligent
exploration of chemical space. In this methodology, an RL agent interacts with a simulated
environment representing molecular structures and chemical properties. The agent generates,
evaluates, and optimizes candidate molecules based on a reward system designed to favor
desirable characteristics such as efficacy, stability, and low toxicity. Key RL algorithms applied
in this domain include Deep Q-Networks (DQN), Policy Gradient Methods, and Actor-Critic
Models, which guide the agent to discover optimal molecular structures efficiently.
This seminar also highlights real-world applications and case studies, demonstrating how RL
can significantly reduce the time and cost of drug discovery while improving precision in
molecular design. The findings indicate that combining RL with computational chemistry and
molecular simulations presents a transformative approach for the pharmaceutical industry,
paving the way for faster, more cost-effective, and accurate drug development.
