Tutorials

Tutorials will be held on the first day of the Congress. You can register for the tutorial after you have registered for the Congress through the registration system.

If you are unable to attend the tutorial after registering, please contact the secretariat in advance. Seating for tutorials is limited, and we would appreciate your cooperation in making sure that as many people as possible who wish to participate in tutorials can do so.

October 23, 2023 (Mon.) 13:00-17:00, 307, Tower Hall Funabori 3F
Free of charge, registration required (capacity of 30 people)

Purpose
As in the bioinformatics field, many open source software (OSS) are used for research in the chemoinformatics field as well as in the bioinformatics field, but on the other hand, there is little systematic and organized information, and its use in actual work requires trial and error. In this section, we will hold a hands-on seminar for chemoinformatics practitioners using RDKit, which is used as a standard OSS tool for chemoinformatics.

Specifically, the following topics will be covered in a hands-on format. 
1. Analysis of compound data extracted from patents related to specific targets
Data preprocessing (normalization, desalting, etc.) chemical space visualization, clustering
2. Utilization of prediction models
Prediction model construction (data preparation, model optimization, model creation)
Validation of compounds using predictive models
3. Virtual screening using public data

Target audienceChemoinformatics intermediate level
Those who are able to build environment using Anaconda, etc. by themselves.
Those who are using or want to use RDKit in practice.
Experienced in Python Programming

Moderator
Kazushi Okawa (Asahi Kasei Pharma Corporation)
Takayuki Serizawa (Daiichi Sankyo Company, Limited)
Kazutoshi Takahashi (Ajinomoto Co., Ltd.)
Natsumi Miyano (Teijin Pharma Limited)

Preparation in advance　
We plan to use Jupyter-notebook on the day of the workshop. We will also provide separate information on the codes and data to be used on the day.

October 23, 2023 (Monday) 13:00-17:00 Tower Hall Funabori 4F 401
Free of charge, registration required (capacity of 30 people)

Purpose
The FMO database (FMODB; (https://drugdesign.riken.jp/FMODB/)), which contains fragment molecular orbital (FMO) calculation results, has been increasing the number of data since its public release in February 2019, with 16816 structures available as of March 23, 2023. Recently, FMO calculations using large-scale data such as MD snapshots using supercomputers such as Fugaku have begun. In addition, FMO analysis and data collection efforts focusing on the dynamic structure as well as on the static structure of drug target proteins are expanding. In this session, an overview of FMODB will be presented for users who are considering starting FMO calculations, and analysis methods for drug discovery research will be introduced using actual registered data. For intermediate users who are already performing FMO calculations, the latest research case studies will be presented. In this process, we will experience how to analyze actual FMO calculations in a tutorial format; we look forward to your participation if you would like to learn more about FMODB and actually analyze the results of FMO calculations.

Moderators
Chizuru Watanabe (RIKEN)　
Daisuke Takaya (Osaka University)
Koichiro Kato (Kyushu University)

Program
13:00-13:05 Introduction
Daisuke Takatani (Osaka University)
13:05-14:35 Introduction of FMO database and how to use the data
Chizuru Watanabe (RIKEN) 
Interaction energies (IFIE/PIEDA) obtained by the FMO method, one of quantum chemical calculations, are expected to be applied to protein-ligand interaction analysis and drug discovery research. Our group has developed the FMODB for storing FMO calculation data, and provides a simple interface to view FMO data such as calculations by members of the FMODD consortium and results calculated by automated pre-processing protocols from our website. This section introduces the functions provided by the FMODB and the status of its development. In addition, analyses using BioStation Viewer, a pre/post-GUI for FMO calculations, and other tools will be presented as examples of how FMODB data can be used.

14:35-14:55 Break

14:55-16:25 Dynamic Mean Interaction Analysis between Ligand and Protein using FMODB
    Daimu Matsumoto (Kyushu University)　
Dynamic Mean Interaction Analysis between Ligand and Protein using FMODB web interface and BioStation Viewer will be presented on the subject of main protease (Mpro) of COVID-19. The latest developments in related research will be introduced, and a dynamic average cpf file will be created from multiple cpf files (files containing various numerical values such as IFIE) obtained from FMO calculations for a series of MD snapshot structures, and analysis using the dynamic average cpf file will be practiced.

16:25-16:35 Summary
Koichiro Kato (Kyushu University)

16:35-17:00 Individual consultation

Preparation in advance
In parallel with the explanations in the tutorial, participants will actually operate the FMO database and BioStation Viewer. If you wish to do so, please pre-install BioStation Viewer, which is available on the FMO Drug Discovery Consortium's website, and download BioStation Viewer from https://fmodd.jp/biostationviewer-dl/ BioStation Viewer can be downloaded from .
The latest version of BioStation Viewer and other programs and data to be used in the tutorial will be posted on the following website approximately one week prior to the tutorial: https://drugdesign.riken.jp/pub/CBI2023 tut/

October 23, 2023 (Monday) 13:00-17:00 Seminar Room, Tower Hall Funabori 4F
Free of charge, registration required (capacity of 35 people)

Purpose
　The utilization of large-scale data of genome information and medical/health information is rapidly progressing in Japan and abroad toward the realization of next-generation health care, and the utilization of genome information and medical information has already become a powerful approach in drug discovery. In order to develop a strategy to realize new medicine and drug discovery by forming a foundation of genome information, medical and health information, and utilizing large-scale data, data science and AI, we will share the latest large-scale medical data and analysis technologies through a datathon in which medical data is actually analyzed. We will also discuss what should be done about the protection of personal information and share the direction we should go in the future.
We look forward to your participation if you are interested in real-world medical data analysis. Please bring your own laptop for analysis on the day.

Prerequisite skills for participant:
You must have intermediate or above programming skills in Python or R.
- R users: data manipulation using Base R or Tidyverse.
- Python users: data manipulation using Pandas or similar packages.

Moderators
Kaoru Mogushi (cBioinformatics Corporation/Integrated Education Organization, Tokyo Medical and Dental University)
Soichi Ogishima (Tohoku University Center for Frontier Medical Sciences/Tohoku Medical Megabank Organization)