Contributing Data

Introduction

This page explains every field you fill in when contributing data through the Submit Data form — what each one means, whether it is required, its allowed values, and a worked example. It is written for community curators reading a paper and entering what it reports.

The golden rule: record only what the paper states. Enter facts the paper explicitly reports for a specific study. Do not infer, average across papers, or fill a field from general knowledge. Anything derivable from an identifier — an organism's lineage, an enzyme's EC number and structure, cross-references — is computed automatically. You supply identity and observations; the system supplies derived annotations (see Data Model).

Field legend

• Required — the submission is rejected without it.
• Optional — provide it only if the paper states it; leave blank otherwise. Never guess.
• Auto / derived — computed by the server from an identifier or the actor type. Do not hand-enter; shown for context only.
• Controlled — the value must come from a fixed list (see Controlled Vocabularies). The form offers these as drop-downs.

Records follow the evidence spine: Reference → Interaction → Actor → Experiment → Observation. The one hard requirement for a submission is at least one interaction; everything else is provided as the paper supports it.

How the form is organised

The form is built around records. A record is one actor — a microorganism, a protein, or an engineered protein variant — and underneath it, one section per plastic that actor was tested against. That mirrors the data model exactly: the database stores one interaction per (actor × plastic) pair, so an enzyme tested on PET and on PCL is two interactions sharing one enzyme.

1. Add a record (Microorganism / Protein · enzyme / Protein variant). It arrives with its first plastic section already open — an actor with no plastic is not yet a claim about anything, so there is nothing to save until you fill one in.
2. Fill in the actor — the identity fields at the top of the card. These describe the organism or protein itself and are shared by every plastic underneath it.
3. Fill in each plastic section. Everything here belongs to that pairing, not to the actor: the plastic, the result, and the per-pairing context fields described below. Use Add a plastic for a second polymer the same actor was tested against.
4. Add experiments under a plastic, and observations under an experiment. One method per experiment; one number per observation.

Two fields take a free-form set of key/value entries — culture_collection_ids on the pairing and additional_conditions on an experiment. Each offers a row editor for the ordinary case and a collapsed Advanced: raw JSON panel underneath for anything a flat key/value pair cannot express. The two views are the same value: editing either updates the other.

1. Reference (paper)

The source publication. Start by entering the DOI — the form looks up the title, authors, year, and citation automatically. Every interaction in your submission attaches to this one reference.

*No single field is individually mandatory here, but a submission must contain at least one interaction. Enter the DOI whenever the paper has one; if the lookup fails (a very new or non-indexed paper), fill the title, authors, and year manually.

2. Interaction

The core unit: one actor acting on one plastic, as reported by this paper. Choose the actor type, then fill the matching actor block plus the study-specific context below.

3. Microorganism actor

Species-level identity and intrinsic biological traits only. Strain, genome, culture-collection, and isolation details are study-specific and go on the Interaction and Sample Source instead.

4. Protein actor

Protein identity only. Provide a name plus at least one identifier so the server can fetch the sequence and derive annotations. EC number, signal peptide, cross-references, and structures are all enrichment-derived — do not enter them.

†Provide at least one of genbank_id, uniprot_accession, or sequence so the protein can be identified and de-duplicated.

5. Protein variant actor

An engineered or otherwise modified version of a parent protein (e.g. a stabilised PETase mutant). Use actor_type = protein_variant. You must identify the parent protein (as in Protein above) and describe the variant. The variant receives its own PlasticDB accession automatically — you do not supply one.

6. Plastic

Reference the plastic by name on the interaction — the server resolves it to the canonical plastic row (creating it if genuinely new). Use the standard polymer abbreviation where one exists.

7. Sample source

Where a microorganism was isolated in this study. Attaches to the interaction (study-specific), not to the species. Fill only what the paper reports.

8. Experiment

One assay run: a single method applied under a set of conditions, hung off the interaction. A paper that probes the same interaction with SEM and weight-loss records two experiments. Only method is required; add each condition the paper reports. Observations, substrate composition, and any enzyme kinetics/properties all attach to the experiment.

9. Substrate composition

For a blend or copolymer substrate, list each constituent and its fraction under the experiment. Skip this entirely for single-material substrates. One row per constituent.

10. Observation

A single measurement an experiment produced. One experiment can yield several (e.g. weight loss at 2, 4, and 6 weeks = three observations). Record the number and its unit; the measurement_type names what was quantified, distinct from the experiment's method (how it was measured).

11. Enzyme kinetics & properties

For enzyme experiments, two optional structured blocks capture biochemistry the paper reports. Enter values only when explicitly stated.

Kinetics

Properties

Worked example

Yoshida et al. (2016) report that Pseudideonella sakaiensis (published as Ideonella sakaiensis, since reclassified) 201-F6, isolated near a PET-recycling site, degrades PET film, and that its purified enzyme PETase does the same. Entered as data, that is two interactions under one reference:

Reference

DOI: 10.1126/science.aad6359   (title/authors/year auto-filled)

Interaction A — the whole organism

actor_type: microorganism
plastic:    PET
result:     positive
strain_designation: 201-F6
microorganism:
    name: Pseudideonella sakaiensis
    tax_id: 1547922
    growth_temp_opt: 30
sample_source:
    environment_type: industrial
    country: Japan
    region: Sakai, Osaka
    sample_description: sediment near a PET-bottle recycling facility
experiments:
  - method: weight_loss
    temperature: 30
    duration: 6      duration_unit: w
    substrate_form: amorphous_film
    sole_carbon_source: true
    culture_medium: mineral salts (no other carbon)
    observations:
      - measurement_type: weight_loss   value: 75   unit: "%"
  - method: sem
    observations:
      - measurement_type: surface_morphology
        text_description: pitting and surface erosion of the film

Interaction B — the purified enzyme

actor_type: protein
plastic:    PET
result:     positive
enzyme_production_context: heterologous
enzyme_preparation: purified
protein:
    name: PETase
    genbank_id: GAP38373.1
    gene: petA
    source_microorganism_name: Pseudideonella sakaiensis
experiments:
  - method: hplc
    temperature: 30      ph: 7.0
    observations:
      - measurement_type: monomer_released   value: 0.13   unit: mM   # MHET/TPA

Note how the organism's degradation (A) and the enzyme's activity (B) are recorded as separate claims — each stands on its own evidence.

Curator checklist

• ☐ One interaction per actor–plastic pair the paper reports (including negatives).
• ☐ DOI entered and the paper metadata looks correct.
• ☐ Correct actor type, and enzyme-only results filed as protein — not as an organism claim.
• ☐ Each organism has a tax ID; each protein has an accession (or a printed sequence).
• ☐ Study-specific facts (strain, isolation source) on the interaction, not the species record.
• ☐ Every reported method is its own experiment; every reported number is its own observation.
• ☐ Controlled fields use listed values only; no invented tokens.
• ☐ Nothing entered that the paper doesn't state, and nothing entered that the system derives (lineage, EC, structures, accessions).
• ☐ Checked the Plastics / Microorganisms / Proteins lists to avoid duplicates.
• ☐ Every record has at least one plastic section filled in — an actor with no plastic is not a claim, and the form will refuse it.
• ☐ Per-pairing context (notes, strain, sample source, enzyme preparation) entered under the plastic it belongs to, not copied blindly across all of them.
• ☐ Kinetics and properties recorded under the experiment they were measured in.
• ☐ For a variant: the parent protein identified by accession, and each mutation's raw notation as printed.