MIAPA Draft RFC

From Evolutionary Interoperability and Outreach
Jump to: navigation, search

We are not yet ready to release an initial draft of MIAPA. However, we have discussed the value of disseminating an initial simple version and requesting feedback on it.

MIAPA Draft

MIAPA strawman draft July 2011

Preamble. The aim of the Minimal Information About a Phylogenetic Analysis (MIAPA) checklist is to facilitate the scientific evaluation and re-use of phylogenetic results. The scope of MIAPA is any report of a historical inference from comparative biological data, such as a phylogenetic tree, an ancestral state, or a date of divergence. The MIAPA checklist has been developed to satisfy critical interests of consumers in evaluating and re-using a result, while minimizing the reporting burden on producers.

  1. Informatics processing context. The record of a phylogenetic analysis will draw on accepted practices to make clear how it is to be processed (formats, languages). Each record will include a globally unique and stable identifier (GUID).
  2. Scientific context. The scientific context for the analysis will be provided, including the identities of responsible experimenters, links to any associated publication, and the purpose of the analysis.
  3. Reusable trees. Phylogenetic results will be represented in an electronic form that can be processed without loss of information (e.g., Newick, NeXML; graphics files are not sufficient).
  4. Identifiable OTUs. The OTU objects represented in electronic files will have identifiers that are externally meaningful (e.g., LSIDs) or, if local names are used, will be linked to such identifiers.
  5. Explicit uncertainty. As inferences, phylogenetic results are estimates, therefore a best practice is to present each result with an appropriate measure of uncertainty. Support for clades or bipartitions, if reported in an associated publication, will be included.
  6. Represented inputs. The record will represent, either explicitly, or by external reference, the data on which the phylogenetic inference is based (typically a character matrix or alignment; sometimes other data such as fossil dates).
  7. Described methods. The method by which the phylogenetic result is inferred from the inputs will be described, ideally by drawing on a controlled vocabulary.

MIAPA strawman draft Oct 2011

The MIAPA workshop at the 2011 TDWG conference resulted in a second, more detailed strawman draft. It was developed independently (though with knowledge) of the one from July, by a different set of participants, using parallel working break-out groups and subsequently reconciling attributes in common.

Stakeholder engagement

  • individual producers of phylogenetic results
  • individual consumers of phylogenetic results
  • aggregators
  • archives (journals, TB, Dryad)
  • Journal editors
  • funding agencies

Compliance

Use of the standard may be enhanced by measuring compliance, and providing rewards for compliance or penalties for non-compliance. We could consider compliance or support of the standard relative to

  • a phylogenetic report
  • an archive that accepts phylogenetic records
  • a software tool that generates phylogenetic results
  • the policies or activities of an institution or group

This idea is inspired by ideas for things like LEED green building standards or CQL compliance standards. A minimal set of points or criteria are required to be achieved to meet a given standard (such as being a LEED-certified building) but with more compliance, a different level is achieved (eg., a LEED-gold building). It allows an achievable but useful minimal standard but creates pressure for meeting higher standards (i.e., in grant applications).