Nkeji 2 / 11

Nnọchite anya Molecular na Structure Chemical: Nkwado na SMILES, InChi na RDKit

Uru:

  • Ikike ịchọpụta ụmụ irighiri ihe ọ bụghị n'aha kama site na nnọchite anya nhazi (SMILES na ụmụ mmadụ, InChI/InChIKey nwere nchekwa data)
  • Ikike ịchọpụta ụlọ ezighi ezi ma ọ bụ ezighi ezi site na ịkọwapụta, kwado ya na ịkọwapụta SMILES ọ bụla sitere na ọgụgụ isi wuru na RDKit.
  • Inwe ike ịghọta na ọnụọgụ deterministic dị ka ịdị arọ molekụla na usoro kwesịrị inweta site na ngwaọrụ dabere na iwu, ọ bụghị site na ụdị asụsụ.

Ọnọdụ mbụ maka iji AI n'enweghị nsogbu na kemịkalụ bụ ide molekul n'asụsụ nke igwe ma mmadụ nwere ike ịghọta. Ị na-ekwu "phenol", AI na-enweta ya nke ọma; mana mgbe ị na-ekwu "acid" enwere ọtụtụ puku ohere. Aha ndị ahụ edoghị anya; ihe nnochite anya n'usoro ziri ezi. N'ime nkeji a, anyị ga-amụta akara abụọ bụ isi na-atụgharị molekul n'ime ederede (SMILES na InChi) yana ngwa na-enyocha ha nke ọma (RDKit). Ebumnuche anyị: inye molecule ahụ AI n'ụzọ ọ na-agaghị aghọtahie ya, na iji ngwaọrụ kwadoro nhazi nke AI mepụtara.

Kedu ihe bụ SILES?

SMILES (Simplified Molecular-Input Line-Entry Sistemu) bụ usoro nke ide molekul n'otu ahịrị ederede. Atọm na-anọchi anya mkpụrụedemede, agbụ site na akara, yana mgbanaka site na ọnụọgụ. Ọmụmaatụ:

  • Ethanol: CCO (carbon-carbon-oxygen; hydrogens doro anya).
  • Benzene: c1ccccc1 (obere "c" na-egosi carbon aromatic; 1's mechie mgbanaka).
  • Aspirin: CC(=O)Oc1ccccc1C(=O)O.
  • Caffeine: Cn1cnc2c1c(=O)n(C)c(=O)n2C.

Ike nke SILES bụ na ọ na-ebu usoro njikọ njikọ n'otu ahịrị; Adịghị ike ya bụ na otu molekul ahụ nwere ike inwe ọtụtụ SMILES bara uru (nke a na-akpọ non-canonicality). Anyị ga-eji RDKit dozie nke a n'otu ntabi anya.

Ndụmọdụ: "SILES canonical" maka molecule bụ otu, ọkọlọtọ mkpoputa maka molekul ahụ. Iji hụ ma ọ bụrụ na ọchị abụọ bụ otu molekul, canonicalize na tụnyere ha; Ha ekwesịghị ịha nhata n'ederede, mana ha kwesịrị ịha nhata ụmụ irighiri ihe.

Kedu ihe bụ InChi?

InChi (International Chemical Identifier) ​​bụ njirimara ọkọlọtọ nke IUPAC mepụtara. Ihe dị iche na SILES bụ na otu molekul na-enye mgbe niile otu InChI; ya bu, o bu ihe omuma na okike. Ọ dị ogologo ma sie ike ịgụ, mana ọ dị mma maka ndakọrịta nchekwa data. A na-eji mkpirisi "InChIKey" (agwa 27) maka ịchọ.

  • Aspirin InChIKey: BSYNRYMUTXBXSQ-UHFFFAOYSA-N.

Iwu: Ndị mmadụ na-asụ SMILES (gụọ), igwe na ọdụ data dakọtara InChi/InChIKey (kpọmkwem). Nye ndị AI ọchị ọchị; Jiri InChiKey kwado na nchekwa data.

RDKit: injin nyocha deterministic

RDKit bụ ọba akwụkwọ cheminformatics mepere emepe. N'adịghị ka ụdị asụsụ, ọ na-adabere na iwu: na-atụgharị SMILES, gbakọọ ịdị arọ molecular, na-emepụta SMILES canonical, na-eweghachi InChI/InChIKey, na-ese molecule ahụ. Yabụ RDKit "na-agbakọ" ihe AI ​​"buru amụma". Nke a bụ obi nke usoro ọrụ: AI na-ewepụta, RDKit na-enyocha.

Usoro nkwenye bụ isi:

si rdkit mbubata Chemfrom rdkit.Chem mbubata Nkọwa, Drawsmiles = "CC(=O)Oc1ccccc1C(=O)O" Chem.MolToSmiles(mol)) ibipụta ("Molecular usoro:", Chem.rdMolDescriptors.CalcMolFormula(mol)) ebipụta (.MolMolFormula(mol)) ebipụta (.MolMolscriptors), okirikiri ("MolMol). 2), "g/mol") ebipụta ("InChIKey:", Chem.MolToInchiKey(mol))

Ọ bụrụ na MolFromSmiles alaghachiteghị nke ọ bụla, AI enyela gị molecule na-adịghị mma; Naanị nke a bụ ịdọ aka ná ntị bara uru nke ukwuu. Ọ bụrụ na ọ dị irè, ịgakwaghị ajụ ihe nlereanya asụsụ maka ịdị arọ molekụla; Ọ bụ n'aka gị deterministically.

Nzọụkwụ site nzọụkwụ: AI + RDKit mmekọrịta

  1. Chọpụta molekul: Nye AI aha, jụọ maka SMILES. Dịka ọmụmaatụ, "chọpụta SMILES canonical maka paracetamol."
  2. Nyochaa: Jiri MolFromSmiles tụgharịa SMILES na-abata. Ọ bụrụ na Ọ nweghị, jụ.
  3. Canonicalize: Weghachite mkpoputa canonical na MolToSmiles; Jiri nke a mgbe niile site ugbu a gawa.
  4. Gbakọọ ọnụọgụgụ: Nweta ịdị arọ molekụla, usoro, ọnụọgụ mgbanaka, ọnụ ọgụgụ nke ndị na-enye onyinye / ndị na-anabata njikọ hydrogen wdg sitere na RDKit, ọ bụghị site na AI.
  5. Idozi ID: Mepụta InChIKey, chọọ na nchekwa data (PubChem), gosi na molekul bụ n'ezie ogige ị chọrọ.

Ndebiri anọ enwere ike iṅomi

1) Arịrịọ SILES (site na aha ruo na nhazi):

Ọrụ: Nye SMILES canonical maka ogige ndị a. Ngwakọta: paracetamol (acetaminophen). Iwu: Nye naanị eriri SMILES na InChIKey, etinyekwala nkọwa ọzọ. Ọ bụrụ na ịmaghị n'aka, dee "ỌRỤGHỊ", emela ya.

2) Arịrịọ nkwado SILES (site na nhazi ruo na njikwa):

Nyochaa SMILES n'okpuru: CC(=O)Nc1ccc(O)cc1Ajụjụ:1) Nke a ọ̀ bụ SMILES bara uru?2) Kedu ihe jikọrọ ya na (aha a na-ahụkarị)?3) Gịnị bụ usoro mkpụrụ ndụ ihe nketa? Rịba ama: M ga-eji RDKit gbakọọ ihe dị arọ nke molekul; Naanị ị na-enye nkọwa nhazi gị.

3) Tụnyere ihe nnọchianya abụọ:

Enwere m ọnụ ọchị abụọ:A) OCCB) CCOTask: Ndị a bụ otu molekul ma ọ bụ dị iche? Dee echiche gị. Mara: Aga m agafe-elele azịza gị site n'ịdepụta ya na RDKit.

4) Nkọwa nhazi (maka ebumnuche mmụta):

SMILES: Cn1cnc2c1c(=O)n(C)c(=O)n2CTask: Gụọ mpempe SMILES a n'otu ibe ma kọwaa ihe akara ọ bụla pụtara (atom, bond, ring, aromaticity) na Turkish larịị. Gwakwa ihe ngwakọta ọ bụ.

Ngwa ngwa adịghị ike / Ike siri ike

Adịghị ike:

Nye ihe onwunwe nke acetaminophen.

"Njirimara" edoghị anya; AI na-ewepụta ọnụọgụ ọnụọgụ site na ịdị arọ molekụla ruo ebe mgbaze, ụfọdụ n'ime ha ga-abụ ihe na-ezighi ezi.

siri ike:

Ngwakọta: acetaminophen.1) Canonical SILES na InChIKey ver.2) Nye usoro molecular.Iwu: ENWELA NUMERIC ụkpụrụ dị ka molecular arọ, agbaze ebe, wdg; Aga m enweta ha na RDKit/PubChem. Naanị nye NJ nrụpụta.

Ọdịiche: Anyị na-eduzi AI n'ihe ọ siri ike na (nhazi njirimara) yana pụọ ​​na ihe ọ na-adịghị ike (ọnụọgụ nnwale).

Ntụnyere echiche

atụmatụ

NA-achị ọchị

InChi / InChiKey

Ịgụ ihe

Elu (ndị mmadụ nwere enyi)

Dị ala (igwe enyi na enyi)

canonicity

Isiokwu ga-agbanwe (chọrọ canonization)

ndammana alụghị di

Ojiji

nkwurịta okwu mmadụ, ịbịaru, ntinye

Dakọtara database, njirimara

stereochemistry

Na-akwado (@ akara)

Nkwado (nke edobere)

inye AI

ezigbo

Kwesịrị ekwesị maka nkwenye

Obere ikpe

Ikpe 1 - Abụọ aha, otu molekul. Otu nwa akwụkwọ chere na "N-acetyl-para-aminophenol" na "paracetamol" bụ ogige dị iche iche ma hazie nnwale abụọ dị iche iche. Mgbe edebere SILES ha na RDKit, ha abụọ tụgharịrị bụrụ CC(=O)Nc1ccc(O)cc1; Ọ bụ otu ihe ahụ molekul. Echefuola: ọkara ụbọchị atụmatụ; uru: enwere ike zere ya site na nlele canonicalization nkeji 2.

Ikpe 2 - Mkpebi SMILES na-ezighi ezi. AI weghachiri CC(=O)Nc1ccc(O)cc1C(maka variant(brackets anaghị emechi) nwa akwụkwọ ahụ gbara MolFromSmiles, ọ laghachiri Ọ dịghị, ozugbo hụrụ njehie ahụ wee rịọ ka emezie SMILES. Na-enweghị nkwenye, usoro ahụ na-adịghị mma ga-agbasa na akaụntụ niile.

Ikpe 3 - arọ molekụla ezighi ezi. YZ kwuru "ihe dị arọ 214.3 g/mol" maka ibuprofen (C13H18O2). Mgbakọ RDKit nyere 206.28 g/mol. Ọdịiche maka 0.1 mol: 21.43 g na YZ, n'ezie 20.63 g. Ọdịiche a nke 3.9% ga-akpaghasị stoichiometry wee wepụta mkpokọta. Ọ wetara mgbako deterministic.

Mmejọ ndị nkịtị

  • Inye molekul n'aha. Synonyms/aha ahia gbagwojuru anya. Na-agụnye SMILES ma ọ bụ InChi mgbe niile.
  • Tụnyere SMILES na-enweghị ịkọwapụta ya. OCC na CCO dị iche na ederede mana otu na ụmụ irighiri ihe.
  • Ịjụ molekụla arọ ahụ na ụdị ire. Nke a bụ ngụkọta nke deterministic; Emere ya site na RDKit ma ọ bụ jiri aka mee ya site na usoro.
  • Ahụghị SMILES na-ezighi ezi. Anaghị elele nloghachi nke MolFromSmiles Ọ dịghị ma na-aga n'ihu na usoro adịghị mma.
  • Na-eleghara stereochemistry anya. Abụọ enantiomers (mirror image isomers) nwere ike igosi mmetụta dị iche iche nke ndu; Mwụpụ akara @/@@ na SMILES.
Nlebara anya: Otu usoro ihe omimi ahụ apụtaghị ihe dị iche iche. C2H6O bụ ma ethanol (CCO) na dimethyl ether (COC). Ịha nhata nke usoro abụghị nha anya njirimara; Jiri InChiKey maka njirimara.

Na nchịkọta

  • Chọpụta ụmụ irighiri ihe site na ndetu nhazi, ọ bụghị aha: SMILES na mmadụ, InChI/InChIKey nwere nchekwa data.
  • RDKit bụ mkpebi siri ike; AI na-ebu amụma, RDKit na-agbakọ ma chọpụta.
  • Jiri MolFromSmiles tụsaa AI ọ bụla ọchị wee lelee Ọ nweghị, wee degharịa ya.
  • Nweta ọnụọgụ dị ka ịdị arọ molekụla na usoro sitere na RDKit, ọ bụghị site na ụdị asụsụ.
  • Ịha nhata n'ụdị anaghị apụta njirimara molekụla; Jiri InChiKey maka njirimara.

Ọrụ ngwa

Họrọ ogige atọ (dịka caffeine, ibuprofen, glycine). Rịọ AI maka SMILES canonical maka nke ọ bụla. Wee dee edemede RDKit (jiri template dị n'elu) wee mepụta: SMILES canonical, usoro ihe omimi, ịdị arọ molecular, InChIKey. Ole n'ime SMILES ndị AI nyere bara uru? Ọ bụrụ na YZ nyekwara ịdị arọ molekụla, gbakọọ ihe dị iche dị ka pasent na uru RDKit. Chee echiche gị n'ime obere tebụl.

ndetu

  • [ ] Amaara m ihe SILES na InChi/InChIKey bụ na mgbe m ga-eji ha.
  • Ana m enyocha SMILES ọ bụla AI nyere site na MolFromSmiles.
  • [ ] Ana m akọwa SMILES tupu atụnyere ha.
  • [ ] Ana m enweta ibu molecular na usoro site na RDKit, ọ bụghị site na ụdị asụsụ.
  • [ ] Eji InChIKey kwadoro njirimara molekul dị na nchekwa data.
  • [ ] Amaara m ọnọdụ ebe stereochemistry dị mkpa.