How Computer systems “See” Molecules | In the direction of Knowledge Science

a pc, Edvard Munch’s The Scream is nothing greater than a grid of pixel values. It has no sense of why swirling traces in a twilight sky convey the agony of a scream. That’s as a result of (trendy digital) computer systems essentially course of solely binary indicators [1,2]; they don’t inherently comprehend the objects and feelings we understand.

To imitate human intelligence, we first want an intermediate type (illustration) to “translate” our sensory world into one thing a pc can deal with. For The Scream, that may imply extracting edges, colours, shapes, and so on. Likewise, in Pure Language Processing (NLP), a pc sees human language as an unstructured stream of symbols that should be became numeric vectors or different structured kinds. Solely then can it start to map uncooked enter to higher-level ideas (i.e., constructing a mannequin).

Human intelligence additionally is determined by inside representations.

In psychology, a illustration refers to an inside psychological image or picture that stands for one thing within the outdoors world [3]. In different phrases, a illustration is how info is encoded within the mind: the symbols we use (phrases, pictures, reminiscences, creative depictions, and so on.) to face for objects and concepts.

Our senses don’t merely put the exterior world straight into our brains; as a substitute, they convert sensory enter into summary neural indicators. For instance, the eyes convert gentle into electrical indicators on the retina, and the ears flip air vibrations into nerve impulses. These neural indicators are the mind’s illustration of the exterior world, which is used to reconstruct our notion of actuality, primarily constructing a “mannequin” in our thoughts.

Between ages one and two, kids enter Piaget’s early preoperational stage [4]. That is when youngsters begin utilizing one factor to characterize one other: a toddler would possibly maintain a banana as much as their ear and babble as if it’s a cellphone, or push a field round pretending it’s a automotive. This type of symbolic play is essential for cognitive growth, as a result of it reveals the kid can transfer past the here-and-now and mission the ideas of their thoughts onto actuality [5].

With out our senses translating bodily indicators into inside codes, we couldn’t understand something [5].

“Rubbish in, rubbish out”. The standard of a illustration units an higher certain on the efficiency of any mannequin constructed on it [6,7].

A lot of the progress in human intelligence has come from bettering how we characterize data [8].

One of many core targets of training is to assist college students type efficient psychological representations of recent data. Seasoned educators use diagrams, animations, analogies and different instruments to current summary ideas in a vivid, relatable approach. Richard Mayer argues that significant studying occurs when learners type a coherent psychological illustration or mannequin of the fabric, quite than simply memorizing disconnected details [8]. In significant studying, new info integrates into present data, permitting college students to switch and apply it in novel conditions.

Nevertheless, in observe, elements like restricted mannequin capability and finite computing assets constrain how complicated our representations might be. Compressing enter knowledge inevitably dangers info loss, noise, and artifacts. So, as step one, growing a “ok” illustration requires balancing a number of key properties:

• It ought to retain the data essential to the duty. (A clear downside definition helps filter out the remainder.)
• It must be as compact as doable: minimizing redundancy and maintaining dimensionality low.
• It ought to separate lessons in characteristic house. Samples from the identical class cluster collectively, whereas these from completely different lessons keep far aside.
• It must be sturdy to enter noise, compression artifacts, and shifts in knowledge modality.
• Invariance. Representations must be invariant to process‑irrelevant modifications (e.g. rotating or translating a picture, or altering its brightness).
• Generalizability.
• Interpretability.
• Transferability.

These limitations on illustration complexity are considerably analogous to the restricted capability of our personal working reminiscence.

Human short-term reminiscence, on common, can solely maintain about 7±2 gadgets directly [9]. When too many unbiased items of data arrive concurrently (past what our cognitive load can deal with), our brains bathroom down. Cognitive psychology analysis reveals that with the appropriate steerage (by adjusting how info is represented), folks can reorganize info to beat this obvious restrict [10,11]. For instance, we will keep in mind a protracted string of digits extra simply by chunking them into significant teams (which is why cellphone numbers are sometimes cut up into shorter blocks).

Now, shifting from The Scream to the microscopic world of molecules, we face the identical problem: how can we translate real-world molecules right into a type that a pc can perceive? With the appropriate illustration, a pc can infer chemical properties or organic features, and in the end map these to increased‑degree ideas (e.g., a drug’s exercise or a molecule’s protein binding). On this article, we’ll discover the frequent strategies that allow computer systems “see” molecules.

Chemical Formulation

Maybe probably the most easy depiction of a molecule is its chemical system, like C₈H₁₀N₄O₂ (caffeine), which tells us there are 8 carbon atoms, 10 hydrogen atoms, 4 nitrogen atoms and a pair of oxygen atoms. Nevertheless, its very simplicity can also be its limitation: a system conveys nothing about how these atoms are linked (the bonding topology), how they’re organized in house, or the place useful teams are positioned. That’s why isomers (like ethanol and dimethyl ether) each share C₂H₆O but differ fully in construction and properties.

Linear String

One other frequent option to characterize molecules is to encode them as a linear string of characters, a format broadly adopted in databases [12,13].

SMILES

Probably the most traditional instance is SMILES (Simplified Molecular Enter Line Entry System) [14], developed by David Weininger within the Eighties. SMILES treats atoms as nodes and bonds as edges, then “flattens” them right into a 1D string by way of a depth‑first traversal, preserving all of the connectivity and ring info. Single, double, triple, and fragrant bonds are denoted by the symbols “-”, “=”, “#”, and “:”, respectively. Numbers are used to mark the beginning and finish of rings, and branches off the primary chain are enclosed in parentheses. (See extra in SMILES – Wikipedia.)

SMILES is easy, intuitive, and compact for storage. Its prolonged syntax helps stereochemistry and isotopes. There’s additionally a wealthy ecosystem of instruments supporting it: most chemistry libraries allow us to convert between SMILES and different customary codecs.

Nevertheless, with out an agreed-upon canonicalization algorithm, the identical molecule might be written in a number of legitimate SMILES kinds. This may doubtlessly result in inconsistencies or “knowledge air pollution”, particularly when merging knowledge from a number of sources.

InChI

One other broadly used string format is InChI (Worldwide Chemical Identifier) [15], launched by IUPAC in 2005, to generate globally standardized, machine-readable, and distinctive molecule identifiers. InChI strings, although longer than SMILES, encode extra particulars in layers (together with atoms and their bond connectivity, tautomeric state, isotopes, stereochemistry, and cost), every with strict guidelines and precedence. (See extra in InChI – Wikipedia.)

As a result of an InChI string can change into very prolonged as a molecule grows extra complicated, it’s usually paired with a 27‑character InChIKey hash [15]. The InChIKeys aren’t human‑pleasant, however they’re ultimate for database indexing and for exchanging molecule identifiers throughout programs.

Molecular Descriptor

Many computational fashions require numeric inputs. In comparison with linear string representations, molecular descriptors flip a molecule’s properties and patterns into a vector of numerical options, delivering passable efficiency in lots of duties [7, 16-18].

Todeschini and Consonni describe the molecular descriptor because the “closing results of a logical and mathematical process, which transforms chemical info encoded inside a symbolic illustration of a molecule right into a helpful quantity or the results of some standardized experiment” [16].

We will consider a set of molecular descriptors as a standardized “bodily examination sheet” for a molecule, asking questions like:

• Does it have a benzene ring?
• What number of carbon atoms does it have?
• What’s the expected octanol-water partition coefficient (LogP)?
• Which useful teams are current?
• What’s its 3D conformation or electron distribution like?
• …

Their solutions can take varied kinds, resembling numerical values, categorical flags, vectors, graph-based constructions, tensors and so on. As a result of each molecule in our dataset is described utilizing the identical set of questions (the identical “bodily examination sheet”), comparisons and mannequin inputs change into easy. And since every characteristic has a transparent that means, descriptors enhance the interpretability of the mannequin.

After all, simply as a bodily examination sheet can’t seize completely every part about an individual’s well being, a finite set of molecular descriptors can by no means seize all elements of a molecule’s chemical and bodily nature. Computing descriptors is usually a non-invertible course of, inevitably resulting in a lack of info, and the outcomes aren’t assured to be distinctive. Due to this fact, there are various kinds of molecular descriptors, every specializing in completely different elements.

1000’s of molecular descriptors have been developed over time (for instance, RDKit [19], CDK [20], Mordred [17], and so on.). They are often broadly categorized by the dimensionality of data they encode (these classes aren’t strict divisions):

• 0D: system‑primarily based properties unbiased of construction (e.g., atom counts or molecular weight).
• 1D: sequence-based properties (e.g., counts of sure useful teams).
• 2D: derived from the 2D topology (e.g., eccentric connectivity index [21]).
• 3D: derived from 3D conformation, capturing geometric or spatial properties (e.g., charged partial floor space [22]).
• 4D and better: these incorporate further dimensions resembling time, ensemble, or environmental elements (e.g., descriptors derived from molecular dynamics simulations, or from quantum chemical calculations like HOMO/LUMO).
• Descriptors obtained from different sources together with experimental measurements.

Molecular fingerprints are a particular type of molecular descriptor that encode substructures right into a fixed-length numerical vector [16]. This table summarizes some generally used molecular fingerprints [23], resembling MACCS [24], which is proven within the determine beneath.

Equally, human fingerprints or product barcodes will also be seen as (or transformed to) fixed-format numerical representations.

Totally different descriptors describe molecules from varied elements, so their contributions to completely different duties naturally range. In a process of predicting the aqueous solubility of drug-like molecules, over 4,000 computed descriptors had been evaluated, however solely about 800 made vital contributions to the prediction [7].

Level Cloud

Typically, we’d like our fashions to study straight from a molecule’s 3D construction. For instance, that is essential after we’re all for how two molecules would possibly work together with one another [25], want to look the doable conformations of a molecule [26], or need to simulate its habits in a sure setting [27].

One easy option to characterize a 3D construction is as some extent cloud of its atoms [28]. In different phrases, some extent cloud is a set of coordinates of the atoms in 3D house. Nevertheless, whereas this illustration reveals which atoms are close to one another, it doesn’t explicitly inform us which pairs of atoms are bonded. Inferring connectivity from interatomic distances (e.g., by way of cutoffs) might be error-prone, and should miss increased‑order chemistry like aromaticity or conjugation. Furthermore, our mannequin should account for modifications of uncooked coordinates attributable to rotation or translation. (Extra on this later.)

Graph

A molecule will also be represented as a graph, the place atoms (nodes) are linked by bonds (edges). Graph representations elegantly deal with rings, branches, and complicated bonding preparations. For instance, in a SMILES string, a benzene ring should be “opened” and denoted by particular symbols, whereas in a graph, it’s merely a cycle of nodes linked in a loop.

Molecules are generally modeled as undirected graphs (since bonds haven’t any inherent course) [29-31]. We will additional “embellish” the graph with further domain-specific data to make the illustration extra interpretable: tagging nodes with atom options (e.g., ingredient kind, cost, aromaticity) and edges with bond properties (e.g., order, size, power). Due to this fact,

• (uniqueness) every distinct molecular construction may correspond to a singular graph, and
• (reversibility) we may reconstruct the unique molecule from its graph illustration.

Chemical reactions primarily contain breaking bonds and forming new ones. Utilizing graphs makes it simpler to trace these modifications. Some response‑prediction fashions encode reactants and merchandise as graphs and infer the transformation by evaluating them [32,33].

Graph Neural Networks (GNNs) can straight course of graphs and study from them. Utilizing molecular graph illustration, these fashions can naturally deal with molecules of arbitrary dimension and topology. In truth, many GNNs have outperformed fashions that solely relied on descriptors or linear strings on many molecular duties [7,30,34].

Usually, when a GNN makes a prediction, we will examine which elements of the graph had been most influential. These “essential bits” incessantly correspond to precise chemical substructures or useful teams. In distinction, if we had been a selected substring of a SMILES, it’s not assured to map neatly to a significant substructure.

A graph doesn’t all the time imply simply the direct bonds connecting atoms. We will assemble completely different sorts of graphs from molecular knowledge relying on our wants, and typically these alternate graphs yield higher outcomes for explicit functions. For instance:

Full graph: Each pair of nodes is linked by an edge. It may introduce redundant connections, however is perhaps used to let a mannequin take into account all pairwise interactions.
Bipartite graph: Nodes are divided into two units, and edges solely join nodes from one set to nodes from the opposite.
Nearest-neighbor graph: Every node is linked solely to its nearest neighbors (in accordance with some criterion), for controlling complexity.

Extensible Graph Representations

We will incorporate chemical guidelines or impose constraints inside molecular graphs. In de novo molecular design, (early) SMILES‑primarily based generative fashions usually produced SMILES strings ended up proposing invalid molecules, as a result of: (1) assembling characters might break SMILES syntax, and (2) even a syntactically appropriate SMILES would possibly encode an unimaginable construction. Graph‑primarily based generative fashions keep away from them by constructing molecules atom by atom and bond by bond (below user-specified chemical guidelines). Graphs additionally allow us to impose constraints: require or forbid particular substructures, implement 3D shapes or chirality, and so forth; thus, to information technology towards legitimate candidates that meet our targets [35,36].

Molecular graphs can even deal with a number of molecules and their interactions (e.g., drug-protein binding, protein-protein interfaces). “Graph-of-graphs” deal with every molecule as its personal graph, then deploy a higher-level mannequin to learn the way they work together [37]. Or, we might merge the molecules into one composite graph, together with all atoms from each companions and add particular (dummy) edges or nodes to mark their contacts [38].

Up to now, we’ve been contemplating the usual graph of bonds (the 2D connectivity), however what if the 3D association issues? Graph representations can definitely be augmented with 3D info: 3D coordinates may very well be hooked up to every node, or distances/angles may very well be added as attributes on the perimeters, to make fashions extra delicate to distinction in 3D configurations. A greater possibility is to make use of fashions like SE(3)-equivariant GNNs, which guarantee their outputs (or key inside options) remodel (or keep invariant) with any rotation or translation of the enter.

In 3D house, the particular Euclidean group SE(3) describes all doable inflexible motions (any mixture of rotations and translations). (It’s typically described as a semidirect product of the rotation group SO(3) with the interpretation group R³.) [28]

After we say a mannequin or a perform has SE(3) invariance, we imply that it offers the identical consequence regardless of how we rotate or translate the enter in 3D. This type of invariance is commonly an important requirement for a lot of molecular modeling duties: a molecule floating in resolution has no mounted reference body (i.e., it will possibly tumble round in house). So, if we predict some property of the molecule (say its binding affinity), that prediction shouldn’t be influenced by the molecule’s orientation or place.

Sequence Representations of Biomacromolecules

We’ve talked largely about small molecules. However organic macromolecules (like proteins, DNA, and RNA) can include 1000’s and even tens of millions of atoms. SMILES or InChI strings change into extraordinarily lengthy and complicated, resulting in the related large computational, storage, and evaluation prices.

This brings us again to the significance of defining the issue: for biomacromolecules, we’re usually not within the exact place of each single atom or the precise bonds between every pair of atoms. As a substitute, we care about higher-level structural patterns and useful modules: like a protein’s amino acid spine and its alpha‑helices or beta‑sheets, which fold into tertiary and quaternary constructions. For DNA and RNA, we might care about nucleotide sequences and motifs.

We describe these organic polymers as sequences of their constructing blocks (i.e., major construction): proteins as chains of amino acids, and DNA/RNA as strings of nucleotides. There are well-established codes for these constructing blocks (outlined by IUPAC/IUBMB): as an example, in DNA, the letters A, C, G, T characterize the bases adenine, cytosine, guanine, and thymine respectively.

Static Embeddings and Pretrained Embeddings

To transform a sequence into numerical vectors, we will use static embeddings: assigning a set vector to every residue (or k-mer fragment). The best static embedding is one-hot encoding (e.g., encode adenine A as [1,0,0,0]), turning a sequence right into a matrix. One other method is to study dense (pretrained) embeddings by leveraging massive databases of sequences. For instance, ProtVec [39] breaks proteins into overlapping 3‑mers and trains a Word2Vec‑like mannequin (generally utilized in NLP) on a big corpus of sequences, assigning every 3-mer a 100D vector. These realized fragment embeddings are proven to seize biochemical and biophysical patterns: fragments with related features or properties cluster nearer within the embedding house.

k-mer fragments (or k-mers) are substrings of size ok extracted from a organic sequence.

Tokens

Impressed by NLP, we will deal with a sequence as if it’s a sentence composed of tokens or phrases (i.e., residues or k-mer fragments), after which feed them into deep language fashions. Educated on large collections of sequences, these fashions study biology’s “grammar” and “semantics” simply as they do in human language.

Transformers can use self‑consideration to seize lengthy‑vary dependencies in sequences; and we primarily use them to study a “language of biology”. (Some) Meta’s ESM sequence of fashions [40-42] educated Transformers on lots of of tens of millions of protein sequences. Equally, DNABERT [43] tokenizes DNA into ok‑mers for BERT coaching on genomic knowledge. These sorts of obtained embeddings have been proven to encapsulate a wealth of organic info. In lots of instances, these embeddings can be utilized straight for varied duties (i.e., switch studying).

Descriptors

In observe, sequence-based fashions usually mix their embeddings with physicochemical properties, statistical options, and different descriptors, resembling the share of every amino acid in a protein, the GC content material of a DNA sequence, or indices like hydrophobicity, polarity, cost, and molecular quantity.

Past the primary classes above, there are another unconventional methods to characterize sequences. Chaos Recreation Illustration (CGR) [44] maps DNA sequences to factors in a 2D airplane, creating distinctive picture patterns for downstream evaluation.

Structural Representations of Biomacromolecules

The complicated construction (of a protein) determines its features and specificities [28]. Merely figuring out the linear sequence of residues is commonly not sufficient to totally perceive a biomolecule’s perform or mechanism (i.e., sequence-structure hole).

Constructions are typically extra conserved than sequences [28, 45]. Two proteins may need very divergent sequences however nonetheless fold into extremely related 3D constructions [46]. Fixing the construction of a biomolecule may give insights that we wouldn’t get simply from the sequence alone.

Granularity and Dimensionality Management

A single biomolecule might include on the order of 10³-10⁵ atoms (or much more). Encoding each atom and bond explicitly into numerical type produces prohibitively high-dimensional, sparse representations.

Including dimensions to the illustration can shortly run into the curse of dimensionality. As we improve the dimensionality of our knowledge, the “house” we’re asking our mannequin to cowl grows exponentially. Knowledge factors change into sparser relative to that house (it’s like having just a few needles in an ever-expanding haystack). This sparsity means a mannequin would possibly want vastly extra coaching examples to search out dependable patterns. In the meantime, the computational price of processing the information usually grows polynomially or worse with dimensionality.

Not each atom is equally essential for the query we care about: we regularly flip to regulate the granularity of our illustration or scale back dimensionality in good methods (such knowledge usually has a lower-dimensional efficient illustration that may describe the system with out (vital) efficiency loss [47]):

• For proteins, every amino acid might be represented by the coordinates of simply its alpha carbon (C_α). For nucleic acids, one would possibly take every nucleotide and characterize it by the place of its phosphate group or by the middle of its base or sugar ring.
• One other instance of managed granularity comes from how AlphaFold [49] represents protein utilizing spine inflexible teams (or frames). Primarily, for every amino acid, a small set of main-chain atoms, sometimes the N, C_α, C (and perhaps O) are handled as a unit. The relative geometry of those atoms is nearly mounted (covalent bond lengths and angles don’t range considerably), in order that unit might be thought-about as a inflexible block. As a substitute of monitoring every atom individually, the mannequin tracks the place and orientation of that total block in house, decreasing the dangers related to extreme levels of freedom [28] (i.e., errors from the inner motion of atoms inside a residue).

• If we’ve got a big set of protein constructions (or a protracted molecular dynamics trajectory), it may be helpful to cluster these conformations into just a few consultant states. That is usually finished when constructing Markov state fashions: by clustering steady states right into a finite set of discrete “metastable” states, we will simplify a posh vitality panorama right into a community of some states linked by transition possibilities.

Many coarse-grained molecular dynamics pressure fields, resembling MARTINI [50] and UNRES [51], have been developed to characterize structural particulars utilizing fewer particles.

• To seize for side-chain results with out modelling all inside atoms or including extreme levels of freedom, a typical method is to characterize every side-chain with a single level, sometimes its middle of mass [52]. Such side-chain centroid fashions are sometimes used at the side of spine fashions.
• The 3Di Alphabet launched by Foldseek [53] defines a 3D interplay “alphabet” of 20 states that describe protein tertiary interactions. Thus, a protein’s 3D construction might be transformed right into a sequence of 20 symbols; and two constructions might be aligned by aligning their 3Di sequences.
• We might spatially crop or concentrate on simply a part of a biomolecule. As an illustration, if we’re finding out how a small drug molecule binds to a protein (say, in a dataset like PDBBind [54], which is stuffed with protein-ligand complexes), we might solely feed the pockets and medicines into our mannequin.
• Combining completely different granularities or modalities of information.

Level Cloud

We may mannequin a biomacromolecule as a large 3D level cloud of each atom (or residue). As famous earlier, the identical limitations apply.

Distance Matrix

A distance matrix data all pairwise distances between sure key atoms (for proteins, generally the C_α of every amino acid), and is inherently invariant to rotation and translation attributable to its symmetric nature. A contact map simplifies this additional by indicating solely which pairs of residues are “shut sufficient” to keep in touch. Nevertheless, each representations lose directional info; so not all structural particulars might be recovered from them alone.

Graph

Equally, identical to we will use graphs for small molecules, we will use graphs for macromolecular constructions [55,56]. As a substitute of atoms, every node would possibly characterize a bigger unit (see Granularity and Dimensionality Management). To enhance interpretability, further data like residue descriptors and identified interplay networks inside a protein, may be included in nodes and edges. Notice that the graph illustration for biomacromolecules inherits lots of the benefits we mentioned for small molecules.

For macromolecules, edges are sometimes pruned to maintain the graph sparse and manageable in dimension: primarily a type of native magnification that focuses on native substructures, whereas far-apart relationships are handled as background context.

Basic dimensionality discount strategies resembling PCA, t-SNE and UMAP are additionally broadly used to research the high-dimensional structural knowledge of macromolecules. Whereas they don’t give us representations for computation in the identical sense because the others we’ve mentioned, they assist mission complicated knowledge into decrease dimensions (e.g., for visualization or insights).

Latent House

After we prepare a mannequin (particularly generative fashions), it usually learns to encode knowledge right into a compressed inside illustration. This inside illustration lives in some house of decrease dimension, often called the latent house. Consider London’s complicated city format, dense and complex, whereas the latent house is sort of a “map” that captures its essence in a simplified type.

Latent areas are normally indirectly interpretable, however we will discover them by seeing how modifications in latent variables map to modifications within the output. In molecular technology, if a mannequin maps molecules right into a latent house, we will take two molecules (say, as two factors in that house) and generate a path between them. Ochiai et. al. [57] did this by taking two identified molecules as endpoints, interpolating between their latent representations, and decoding the intermediate factors. The consequence was a set of recent molecules that blended options of each originals: hybrids that may have blended properties of the 2.

---

—— About Writer ——

Tianyuan Zheng
[email protected] | [email protected]
Computational Biology, Bioinformatics, Synthetic Intelligence

Division of Pc Science and Know-how
Division of Utilized Arithmetic and Theoretical Physics
College of Cambridge

---

Reference

1. Patterson DA, Hennessy JL. Pc group and design ARM version: the {hardware} software program interface. Morgan kaufmann; 2016 Could 6.
2. Harris S, Harris D. Digital Design and Pc Structure, RISC-V Version. Morgan Kaufmann; 2021 Jul 12.
3. Kosslyn SM, Koenig O. Moist thoughts: The brand new cognitive neuroscience. Simon and Schuster; 1992.
4. Piaget J, Prepare dinner M. The origins of intelligence in kids. New York: Worldwide universities press; 1952.
5. Bergen D. The function of faux play in kids’s cognitive growth. Early Childhood Analysis & Follow. 2002;4(1):n1.
6. Bengio Y, Courville A, Vincent P. Illustration studying: A overview and new views. IEEE transactions on sample evaluation and machine intelligence. 2013 Mar 7;35(8):1798-828.
7. Zheng T, Mitchell JB, Dobson S. Revisiting the applying of machine studying approaches in predicting aqueous solubility. ACS omega. 2024 Jul 31;9(32):35209-22.
8. Mayer RE. Multimedia studying. In Psychology of studying and motivation 2002 Jan 1 (Vol. 41, pp. 85-139). Tutorial Press.
9. Miller GA. The magical quantity seven, plus or minus two: Some limits on our capability for processing info. Psychological overview. 1956 Mar;63(2):81.
10. Chase WG, Simon HA. Notion in chess. Cognitive psychology. 1973 Jan 1;4(1):55-81.
11. Simon HA. How Large Is a Chunk? By combining knowledge from a number of experiments, a fundamental human reminiscence unit might be recognized and measured. Science. 1974 Feb 8;183(4124):482-8.
12. Kim S, Chen J, Cheng T, Gindulyte A, He J, He S, Li Q, Shoemaker BA, Thiessen PA, Yu B, Zaslavsky L. PubChem 2025 replace. Nucleic acids analysis. 2025 Jan 6;53(D1):D1516-25.
13. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE. The protein knowledge financial institution. Nucleic acids analysis. 2000 Jan 1;28(1):235-42.
14. Weininger D. SMILES, a chemical language and data system. 1. Introduction to methodology and encoding guidelines. Journal of chemical info and pc sciences. 1988 Feb 1;28(1):31-6.
15. Heller S, McNaught A, Stein S, Tchekhovskoi D, Pletnev I. InChI-the worldwide chemical construction identifier customary. Journal of cheminformatics. 2013 Jan 24;5(1):7.
16. Todeschini R, Consonni V. Molecular descriptors for chemoinformatics: quantity I: alphabetical itemizing/quantity II: appendices, references. John Wiley & Sons; 2009 Oct 30.
17. Moriwaki H, Tian YS, Kawashita N, Takagi T. Mordred: a molecular descriptor calculator. Journal of cheminformatics. 2018 Feb 6;10(1):4.
18. Jaganathan Ok, Tayara H, Chong KT. An explainable supervised machine studying mannequin for predicting respiratory toxicity of chemical substances utilizing optimum molecular descriptors. Pharmaceutics. 2022 Apr 11;14(4):832.
19. RDKit: Open-source cheminformatics. https://www.rdkit.org
20. Willighagen EL, Mayfield JW, Alvarsson J, Berg A, Carlsson L, Jeliazkova N, Kuhn S, Pluskal T, Rojas-Chertó M, Spjuth O, Torrance G. The Chemistry Growth Equipment (CDK) v2. 0: atom typing, depiction, molecular formulation, and substructure looking. Journal of cheminformatics. 2017 Jun 6;9(1):33.
21. Sharma V, Goswami R, Madan AK. Eccentric connectivity index: A novel extremely discriminating topological descriptor for construction− property and construction− exercise research. Journal of chemical info and pc sciences. 1997 Mar 24;37(2):273-82.
22. Stanton DT, Jurs PC. Growth and use of charged partial floor space structural descriptors in computer-assisted quantitative structure-property relationship research. Analytical Chemistry. 1990 Nov 1;62(21):2323-9.
23. Boldini D, Ballabio D, Consonni V, Todeschini R, Grisoni F, Sieber SA. Effectiveness of molecular fingerprints for exploring the chemical house of pure merchandise. Journal of Cheminformatics. 2024 Mar 25;16(1):35.
24. Durant JL, Leland BA, Henry DR, Nourse JG. Reoptimization of MDL keys to be used in drug discovery. Journal of chemical info and pc sciences. 2002 Nov 25;42(6):1273-80.
25. Kitchen DB, Decornez H, Furr JR, Bajorath J. Docking and scoring in digital screening for drug discovery: strategies and functions. Nature opinions Drug discovery. 2004 Nov 1;3(11):935-49.
26. Friedrich NO, Meyder A, de Bruyn Kops C, Sommer Ok, Flachsenberg F, Rarey M, Kirchmair J. Excessive-quality dataset of protein-bound ligand conformations and its software to benchmarking conformer ensemble mills. Journal of chemical info and modeling. 2017 Mar 27;57(3):529-39.
27. Karplus M, McCammon JA. Molecular dynamics simulations of biomolecules. Nature structural biology. 2002 Sep 1;9(9):646-52.
28. Zheng T, Rondina A, Micklem G, Lio P. Challenges and Tips in Deep Generative Protein Design: 4 Case Research.
29. Duvenaud DK, Maclaurin D, Iparraguirre J, Bombarell R, Hirzel T, Aspuru-Guzik A, Adams RP. Convolutional networks on graphs for studying molecular fingerprints. Advances in neural info processing programs. 2015;28.
30. Gilmer J, Schoenholz SS, Riley PF, Vinyals O, Dahl GE. Neural message passing for quantum chemistry. InInternational convention on machine studying 2017 Jul 17 (pp. 1263-1272). Pmlr.
31. Wu Z, Pan S, Chen F, Lengthy G, Zhang C, Yu PS. A complete survey on graph neural networks. IEEE transactions on neural networks and studying programs. 2020 Mar 24;32(1):4-24.
32. Jin W, Coley C, Barzilay R, Jaakkola T. Predicting natural response outcomes with weisfeiler-lehman community. Advances in neural info processing programs. 2017;30.
33. Shi C, Xu M, Guo H, Zhang M, Tang J. A graph to graphs framework for retrosynthesis prediction. InInternational convention on machine studying 2020 Nov 21 (pp. 8818-8827). PMLR.
34. Wu Z, Ramsundar B, Feinberg EN, Gomes J, Geniesse C, Pappu AS, Leswing Ok, Pande V. MoleculeNet: a benchmark for molecular machine studying. Chemical science. 2018;9(2):513-30.
35. Lim J, Ryu S, Kim JW, Kim WY. Molecular generative mannequin primarily based on conditional variational autoencoder for de novo molecular design. Journal of cheminformatics. 2018 Jul 11;10(1):31.
36. Maziarka Ł, Pocha A, Kaczmarczyk J, Rataj Ok, Danel T, Warchoł M. Mol-CycleGAN: a generative mannequin for molecular optimization. Journal of Cheminformatics. 2020 Jan 8;12(1):2.
37. Wang H, Lian D, Zhang Y, Qin L, Lin X. Gognn: Graph of graphs neural community for predicting structured entity interactions. arXiv preprint arXiv:2005.05537. 2020 Could 12.
38. Jiang D, Hsieh CY, Wu Z, Kang Y, Wang J, Wang E, Liao B, Shen C, Xu L, Wu J, Cao D. InteractionGraphNet: a novel and environment friendly deep graph illustration studying framework for correct protein–ligand interplay predictions. Journal of medicinal chemistry. 2021 Dec 8;64(24):18209-32.
39. Asgari E, Mofrad MR. Steady distributed illustration of organic sequences for deep proteomics and genomics. PloS one. 2015 Nov 10;10(11):e0141287.
40. Rives A, Meier J, Sercu T, Goyal S, Lin Z, Liu J, Guo D, Ott M, Zitnick CL, Ma J, Fergus R. Organic construction and performance emerge from scaling unsupervised studying to 250 million protein sequences. Proceedings of the Nationwide Academy of Sciences. 2021 Apr 13;118(15):e2016239118.
41. Rao R, Meier J, Sercu T, Ovchinnikov S, Rives A. Transformer protein language fashions are unsupervised construction learners. Biorxiv. 2020 Dec 15:2020-12.
42. Rao RM, Liu J, Verkuil R, Meier J, Canny J, Abbeel P, Sercu T, Rives A. MSA transformer. InInternational convention on machine studying 2021 Jul 1 (pp. 8844-8856). PMLR.
43. Ji Y, Zhou Z, Liu H, Davuluri RV. DNABERT: pre-trained Bidirectional Encoder Representations from Transformers mannequin for DNA-language in genome. Bioinformatics. 2021 Aug 1;37(15):2112-20.
44. Jeffrey HJ. Chaos recreation illustration of gene construction. Nucleic acids analysis. 1990 Apr 25;18(8):2163-70.
45. Illergård Ok, Ardell DH, Elofsson A. Construction is three to 10 occasions extra conserved than sequence—a research of structural response in protein cores. Proteins: Construction, Perform, and Bioinformatics. 2009 Nov 15;77(3):499-508.
46. Chothia C, Lesk AM. The relation between the divergence of sequence and construction in proteins. The EMBO journal. 1986 Apr 1;5(4):823-6.
47. Roel-Touris J, Don CG, V. Honorato R, Rodrigues JP, Bonvin AM. Much less is extra: coarse-grained integrative modeling of huge biomolecular assemblies with HADDOCK. Journal of chemical concept and computation. 2019 Sep 20;15(11):6358-67.
48. Duong VT, Diessner EM, Grazioli G, Martin RW, Butts CT. Neural Upscaling from Residue-Stage Protein Construction Networks to Atomistic Constructions. Biomolecules. 2021 Nov 30;11(12):1788.
49. Jumper J, Evans R, Pritzel A, Inexperienced T, Figurnov M, Ronneberger O, Tunyasuvunakool Ok, Bates R, Žídek A, Potapenko A, Bridgland A. Extremely correct protein construction prediction with AlphaFold. nature. 2021 Aug 26;596(7873):583-9.
50. Marrink SJ, Risselada HJ, Yefimov S, Tieleman DP, De Vries AH. The MARTINI pressure area: coarse grained mannequin for biomolecular simulations. The journal of bodily chemistry B. 2007 Jul 12;111(27):7812-24.
51. Liwo A, Baranowski M, Czaplewski C, Gołaś E, He Y, Jagieła D, Krupa P, Maciejczyk M, Makowski M, Mozolewska MA, Niadzvedtski A. A unified coarse-grained mannequin of organic macromolecules primarily based on mean-field multipole–multipole interactions. Journal of molecular modeling. 2014 Aug;20(8):2306.
52. Cao F, von Bülow S, Tesei G, Lindorff‐Larsen Ok. A rough‐grained mannequin for disordered and multi‐area proteins. Protein Science. 2024 Nov;33(11):e5172.
53. Van Kempen M, Kim SS, Tumescheit C, Mirdita M, Lee J, Gilchrist CL, Söding J, Steinegger M. Quick and correct protein construction search with Foldseek. Nature biotechnology. 2024 Feb;42(2):243-6.
54. Wang R, Fang X, Lu Y, Yang CY, Wang S. The PDBbind database: methodologies and updates. Journal of medicinal chemistry. 2005 Jun 16;48(12):4111-9.
55. Ingraham J, Garg V, Barzilay R, Jaakkola T. Generative fashions for graph-based protein design. Advances in neural info processing programs. 2019;32.
56. Jing B, Eismann S, Suriana P, Townshend RJ, Dror R. Studying from protein construction with geometric vector perceptrons. arXiv preprint arXiv:2009.01411. 2020 Sep 3.
57. Ochiai T, Inukai T, Akiyama M, Furui Ok, Ohue M, Matsumori N, Inuki S, Uesugi M, Sunazuka T, Kikuchi Ok, Kakeya H. Variational autoencoder-based chemical latent house for big molecular constructions with 3D complexity. Communications Chemistry. 2023 Nov 16;6(1):249.