SMILES 😊
Few days after when i started working on this i though it's not the most productive topic to work rich now because every thing i want to do can be already done through rdkit so it will be wast of time and efforts.
Â
Dev Blog
Hello SMILES 🤗
When i strarted working on this porject i barely understood SMILES notation and directly started hunting all the research paper on the internet. There was only one research paper which used to show up everytime and i did't understood any thing but the word DFS caught my eyes it's abbreviated as Depth First Search it's basically a graph algorithm e.x google maps shows the closes route
So eventually after a short while i started learning BFS insted of DFS 🤪, from youtube channel (Codeing Train & other channels). And as always excitment was dead after completing 50% of the tutorial. And At the time i was exploring ChatGTP to write abstract and exploring ML Algorithm to generate code and I fire few random input which i think going to change my life
🤨 Naive Searches
> Can you code me SMILES BFS Algorithm in p5.js
> Protein Homology code in Julia
> how to get a binding site of proteins in js
> Code me a molecular docking scoring function in js
And it gave me a reasonably valid code, i was happy with it.
function scoreDocking(ligand, receptor) {
// Calculate distance between ligand and receptor atoms
// and find the minimum distance for each ligand atom
let distances = [];
for (let i = 0; i < ligand.atoms.length; i++) {
let minDistance = Number.MAX_VALUE;
for (let j = 0; j < receptor.atoms.length; j++) {
let distance = ligand.atoms[i].distance(receptor.atoms[j]);
if (distance < minDistance) {
minDistance = distance;
}
}
distances.push(minDistance);
}
// Calculate the docking score as the sum of the minimum distances
let dockingScore = distances.reduce((sum, distance) => sum + distance, 0);
return dockingScore;
}
As i was quite happy with the reply by ChatGPT, so when i told Ankita mam about designing synthetic scheme over this, she spoke word PARSE and suddenly it struc to me that to understand smile is called parse like JSON.parse(), that made me search any parsing DFS Algorithm for SMILES and i finally found one more hope a ready made code of SMILES in rust (rust is a programming language just like julia), this article (opens in a new tab) now today i am here writing this blog and basic data structure.
interface Mol {
atoms?: Vec<Atom>;
bonds?: Vec<Vec<Bond>>;
}
interface Atom {
element?: Element,
aromatic?:bool,
isotope?:Option<u16>,
hcount?:Option<i8>,
charge?:Option<i8>,
parity?:Option<parity>,
map?:Option<u8>
}
interface Bond{
tid?:Number,
style?:Option<Style>,
}
enum Style{
Single,
Double,
Triple,
Quadruple,
Up,
Down,
Aromatic
}