observation

Deep research

Generated by just research-trait; source: research/traits/upper/observation-deep-research-falcon.md

# Curation-focused report: microbial trait **observation**

**Target:** `METPO:1001000`  
**Category:** UPPER | **Kind:** CLASS | **Mapping:** REVIEWED

## Executive assessment

`METPO:1001000` should be modeled as an **investigation and measurement context**, not as a microbial phenotype, physiological capacity, pathway, or environmental preference. It connects an evaluant or specimen, its biological and environmental context, the assay and protocol used, the resulting data item, and any later interpretation. This reading is consistent with the supplied definition and with the Ontology for Biomedical Investigations (OBI), where an assay is a planned process whose output is typically a data item, while a conclusion is a separate information entity based on the generated data. (bandrowski2016theontologyfor pages 11-13, bandrowski2016theontologyfor pages 8-9)

Consequently, the most defensible TraitMech graph is a **measurement-provenance graph**. Genes, proteins, enzymes, pathways, metabolites, and electron donors or acceptors belong beneath a *specific observed microbial trait*—for example nitrate reduction, growth on glucose, or antimicrobial resistance—not directly beneath generic “observation.” Adding them to this upper class without a specified evaluant and assay would imply biological causality that the term does not express.

## 1. Scope and boundaries

### Positive scope

An observation context should capture:

1. **What was evaluated:** organism, strain, community, culture, specimen, or environmental sample.
2. **Under what conditions:** medium, temperature, pH, oxygen regime, exposure, location, collection time, host state, and other relevant biological or environmental metadata.
3. **How it was evaluated:** assay, protocol, instrument, reagents, sampling, preservation, extraction, sequencing, imaging, or other measurement processes.
4. **What was recorded:** qualitative call, numerical measurement, image-derived feature, taxonomic profile, sequence-derived feature, or another data item.
5. **How reliability was assessed:** controls, reference materials, replicates, calibration, detection limits, and quality-control data.
6. **What was inferred later:** a conclusion or phenotype assertion supported by—but not identical to—the observation datum.

OBI’s input–process–output pattern directly supports this interpretation: assays take material inputs and generate specified data outputs; protocols and study designs guide procedures; and conclusions based on data are distinguished from the data generated during study execution. (bandrowski2016theontologyfor pages 11-13, bandrowski2016theontologyfor pages 8-9, bandrowski2016theontologyfor pages 9-11)

### Boundary cases

- **Observation versus assay:** the assay is the planned measurement process; observation is the broader recording context that may include assay, evaluant, conditions, and output.
- **Observation versus data item:** a datum is the information output, not the entire context that generated it.
- **Observation versus conclusion:** “OD600 = 0.42” or “growth detected” is an observation/data output; “the strain can grow anaerobically on nitrate” is an interpreted trait assertion requiring stated conditions and decision criteria.
- **Observation versus phenotype:** phenotype is the organismal quality or disposition being estimated. The observation is evidence about that phenotype.
- **Observation versus observational study:** the target term is not restricted to epidemiological observational designs; it can cover laboratory, field, sequencing, imaging, and computationally derived measurement contexts.
- **Observation versus environmental preference:** temperature, pH, salinity, oxygen, nutrients, and host factors contextualize a measurement. They become causal biological factors only in a graph for a specified microbial response.
- **Observation versus in silico prediction:** a genome-scale model or classifier output may be recorded as a prediction, but it should not be represented as a direct experimental observation unless separately validated.

## 2. Candidate nodes grouped by type

### Core investigation entities

- **observation** — `METPO:1001000`
- assay — OBI candidate; verify the exact current OBI CURIE before YAML insertion
- investigation
- study design
- protocol / plan specification
- instrument or measurement device
- operator or laboratory
- computational analysis process
- conclusion based on data

### Material and biological entities

- evaluant
- specimen
- microbial isolate
- microbial strain
- microbial community
- culture
- environmental sample
- reference material

Showing the first 60 of 276 lines of findings; the linked file also carries the run's front matter and the prompt it was given — read the full report.