Here's a new post on my first encounter with building a simple deep learning model on manually-compiled adverse drug reactions data (thanks to @baoilleach for feedback) - https://jhylin.github.io/Data_in_life_blog/posts/22_Simple_dnn_adrs/2_ADR_regressor.html

Notes re. data - https://jhylin.github.io/Data_in_life_blog/posts/22_Simple_dnn_adrs/1_ADR_data.html

#PyTorch #RDKit #ChEMBL #embeddings #cheminformatics

At the ChEBI 2.0 workshop, Muhammad Arsalan is presenting how ChEBI is using the Bioregistry to standardize its cross-references, generate URLs on their front-end, and more

#chembl #ebi #chebi #sssom #cheminformatics

An update on an older post looking at saving a relatively large csv file (although may not be considered large by some) as a Parquet file first (to be followed by 3 other smaller posts later detailing the use of Polars with scikit-learn without using Pandas at all)

https://jhylin.github.io/Data_in_life_blog/posts/21_ML1-1-1_Small_mols_in_chembl_update/ML1-1-1_chembl_cpds_parquet_new.html

#Scikit_Learn #Polars #parquet #Python #ChEMBL #Cheminformatics

Here are some snapshots from the #ChEMBL symposium! Dr. Samantha attended & delivered a wonderful talk about the #SemanticWeb! You can find the slides right here ➡️ https://zenodo.org/records/13882075

#PSDI #event #semantics #semanticweb

Don't forget to catch Dr. Samantha's talk about the topic: The Semantic Web is dead, Long live the Semantic Web - The future of Semantics in the Physical Sciences at 14:00 BST at the @chembl symposium. Join virtually using the link in the agenda: https://t.ly/H20xH

#PSDI #CheMBL#event #talk #semantics2024

Dr. Samantha Pearman-Kanza, is speaking at the @chembl 15 year symposium!

In 2024 European Bioinformatics Institute | @emblebi celebrated the 15th anniversary of the first public release of the #ChEMBL database as well as the 10th anniversary of #SureChEMBL.

oh, that's new for me (well, I haven't been using Google a lot lately)... when searching for #ChEMBL and #RDF it actually first lists a few datasets.

Did anyone else see this too? I have it in Chrome/Brave but not in Firefox nor Falkon. Is this because Google knows I like facts (and likely data)?

somewhere in the next months I am going to try to repeat this: https://github.com/egonw/chembl.rdf #chembl #cheminformatics #rdf

I have finally grown more trees leading to this new post on boosted trees - re. chaining Scikit-mol's transformers along with AdaBoost and XGBoost via Scikit_learn's interface and pipelines
https://jhylin.github.io/Data_in_life_blog/posts/19_ML2-3_Boosted_trees/1_adaboost_xgb.html
#cheminformatics #chembl #rdkit #python #ml #xgboost #adaboost #sklearn #scikit_mol

Blog post on "Every ChEMBL everywhere, all at once"
https://baoilleach.blogspot.com/2024/06/every-chembl-everywhere-all-at-once.html
#chembl #cheminformatics

In an attempt to complete the random forest (RF) series, here's another follow-up post on RF classifier with more on imbalanced dataset - https://jhylin.github.io/Data_in_life_blog/posts/17_ML2-2_Random_forest/2_random_forest_classifier.html

#cheminformatics #ml #rf #chembl #chembl_downloader #scikit_mol #rdkit #Scikit_Learn #ghostml #Python

A follow-up on the decision tree series leading to a random forest this time with details on model building, imbalanced dataset, feature importances & hyperparameter tuning - https://jhylin.github.io/Data_in_life_blog/posts/17_ML2-2_Random_forest/1_random_forest.html

Jupyter notebook link: https://github.com/jhylin/ML2-2_random_forest/blob/main/1_random_forest.ipynb

Post updated to show a different max_features used for regression task (thanks @dr_greg_landrum for pointing this out)

#ml #randomforest #scikitlearn #pandas #seaborn #matplotlib #python #cheminformatics #chembl #drugdiscovery

okay, the `curl` command is not correct yet (after shopping/dinner), but the "Run" and "Edit" links are now working for all SPARQL endpoints :) https://bigcat-um.github.io/PRA3006-SPARQL/wikipathways.html #wikidata #wikipathways #chembl #AOPWiki

my lightning talk from the #RDKitUGM2023 is now on YouTube - all about making your work that uses datasets derived from ChEMBL more reproducible

📺 Video: https://youtu.be/PY-xaoRoSOY?list=PLugOo5eIVY3ExzpyKll6GGz4FRgBD2qzN&t=28

📜 Slides: https://bit.ly/cth-rdkit-ugm-2023

🤖 Code/Docs: https://github.com/cthoyt/chembl-downloader

Get started with: pip install chembl-downloader

#cheminformatics #chemoinformatics #chembl #pubchem

Shiny app in R - https://lnkd.in/gGwVYQ2r - this post walks through the process of making a simple Shiny app (without the help of any LLMs). #shiny #rstats #rladies #chembl #cheminformatics

I have my instruction on how to download the #pubChem + #ChEMBL + #NCI60 database.

This goes directly into #neo4j and will only take an hour instead of the weeks my computer crunched to create the database

https://medium.com/p/d9ee9779dfbe

It's always a bit awkward for me to post the first official post on any new platform (my day 3 on 🐘), so here it is, my latest post from my humble portfolio blog on data science, drug discovery, pharmaceuticals and their related data - Re-training and re-evaluation of machine learning model with scikit-learn - ML series 3 on "Small molecules in ChEMBL database" - https://jhylin.github.io/Data_in_life_blog/posts/11_ML3_Small_molecules_in_ChEMBL_database/ML3_chembl_cpds.html #machinelearning #scikitlearn #chembl #cheminformatics #python #pyladies