{
  "field_id": "F037",
  "field": "computational biology",
  "directions": [
    {
      "direction_id": "G0",
      "name": "Small-molecule design and optimization",
      "scope": "Generate or modify candidate small molecules to satisfy design constraints such as protein targets, structural exemplars, or properties. Encompasses both target-guided and non-target-led design."
    },
    {
      "direction_id": "G1",
      "name": "Protein and antibody design",
      "scope": "Propose protein sequences and candidate antibodies with desired structures, properties, or binding capabilities. Includes the full range of antibody sequence, structure, and antigen-interface design."
    },
    {
      "direction_id": "G2",
      "name": "Sequence-based microbial identification",
      "scope": "Identify origins, pathogens, hosts, or types from microbial or viral sequence evidence."
    },
    {
      "direction_id": "G3",
      "name": "Biological interaction characterization",
      "scope": "Infer relationships between biological entities or characterize where binding occurs and how strong it is. Collectively encompasses unknown network edges, protein–small-molecule binding strength, and localization of protein binding residues; an individual study need not produce all of these outputs."
    },
    {
      "direction_id": "G4",
      "name": "Protein structure inference from sequence",
      "scope": "Infer protein structural features or folds from given amino-acid sequences, and improve the inputs and combinations of predictors used for that inference."
    },
    {
      "direction_id": "G5",
      "name": "Genomic regulatory modeling",
      "scope": "Identify regulatory sequence features and model their relationships to epigenetic states or gene expression."
    },
    {
      "direction_id": "G6",
      "name": "Functional and activity annotation of biological entities",
      "scope": "Assign attributes such as function or activity to existing genes, gene sets, proteins, or peptides. Encompasses entity annotation using network evidence and protein and peptide property prediction using sequence evidence."
    },
    {
      "direction_id": "G7",
      "name": "Genomic and transcriptomic marker discovery",
      "scope": "Select compact sets of variants or genes indicative of phenotypes, cell identity, or biological states, including research on the stability of that selection."
    },
    {
      "direction_id": "G8",
      "name": "Single-cell representation and integration",
      "scope": "Recover reusable latent cell structure from single-cell molecular profiles and relate datasets, cell populations, or modalities."
    },
    {
      "direction_id": "G9",
      "name": "Small-molecule property prediction",
      "scope": "Learn chemical representations or similarities of existing small molecules to predict their properties."
    },
    {
      "direction_id": "G10",
      "name": "Molecular prediction of treatment response",
      "scope": "Use drug, genomic, or longitudinal molecular information to predict treatment effectiveness in cellular or patient contexts, rather than molecular binding alone."
    }
  ]
}