Phenoscape use cases

For South Dakota

UI functionality related to identifying and mapping evolutionary changes based on mutant phenotypes

Identify evolutionary phenotypes that match an input phenotype

Motivation

Example

Input

• A phenotype description - e.g. input some entity and/or some quality.

Output

• A list of annotations from the database which match that entity and quality, with taxon.

Full List

1. Identify zebrafish candidate genes for an evolutionary phenotype (ranked #1 by TJV and PM)

Motivation

To obtain one or more candidate genes for an evolutionary change in phenotype, one would like to know which genes, when perturbed in a model organism, give rise to a "similar" phenotypic difference between wildtype and mutant genotypes.

Example

User observes evolutionary variation among fishes in the size of the ceratobranchial 5 bone. User queries Phenoscape for matching mutant zebrafish phenotypes using Entity = Ceratobranchial 5 [from TAO] and all qualities pertaining to attribute ‘size’ [from the PATO]. The response is a list of zebrafish mutants and their phenotypes, along with the associated genes, and possibly gene expression images, in this case for genes such as sox9a, since there is a size reduction in ceratobranchial 5 in the mutant line sox9ahi1134.

Input

A phenotype search specification: entity (TAO), quality (PATO). Option to match exactly or to match descendant terms (as in quality "size", above). [Also, option to match ancestral terms a node or two up will be frequently desired. E.g. if a search yields no phenotype matches to "ceratobranchial 5 bone", the user may want to search "ceratobranchial bone" (Is_a parent, one node up), "ceratobranchial 5 cartilage" (develops_from relationship), "Pharyngeal arch skeleton (Part_of relationship). even pharyngeal arch (more nodes up). User will want to visualize the ontology simultaneously in order to make decisions about what nodes might be informative/how to proceed in a search. --Pmabee@usd.edu 15:50, 18 August 2008 (EDT)]

Output

A list of ZFIN mutant identifiers with matching phenotype(s) of each, associated gene(s), perhaps gene expression images (or links to). [Needs analysis workshop identified "Downloadable reports of query results (e.g. candidate genes; comparative phenotypic data)" as a priority --Pmabee@usd.edu 18:15, 18 August 2008 (EDT)]

Use cases related to identifying and mapping evolutionary changes based on mutant phenotypes

Identify evolutionary changes that match the phenotype of a zebrafish mutant (ranked #2 by TJV & PM; see two other similar mapping use cases, also ranked #2)

Motivation

It would be of interest to identify phenotypic variation among wild species that could plausibly arise from changes in the same gene that is perturbed in particular zebrafish mutant.

Example

User observes a reduction in number of branchiostegal rays in a zebrafish mutant for the gene endothelin-1. The user wants to know whether branchiostegal ray number is variable among fish species, and if so, what is the pattern of change across fish evolution. User queries Phenoscape using Entity = Branchiostegal rays [from TAO] and all qualities pertaining to attribute ‘count’ [from the PATO]. Phenoscape returns a list [just a list?--Tjvision 12:40, 15 August 2008 (EDT)] of taxa and phenotypes. The user would see that all cypriniforms, including zebrafish, have three branchiostegal rays, but other fishes, including close ostariophysan relatives, have higher and lower numbers. [Better to be able to return (default) phylogeny with mapped character state changes --Pmabee@usd.edu 16:00, 18 August 2008 (EDT)]

Input

A phenotype search specification: entity, quality, etc. Option to match exactly or to match descendant terms (as in quality "count", above). [ Also, option to match ancestral terms a node or two up as per above use case --Pmabee@usd.edu 16:02, 18 August 2008 (EDT)]

Output

A list of matching evolutionary phenotypes and the taxon for each. [A list is too crude - the phenotypes should be phylogenetically mapped as in the use case below --Tjvision 12:40, 15 August 2008 (EDT)]

Map changes of an evolutionary character on a phylogeny (ranked #2 by TJV & PM; see two other similar mapping use cases, also ranked #2)

Motivation

After observing phenotypic variation among species in a given trait (as in use case above), the user wants to know what the pattern of evolution of this trait has been.

Example

The user observes variation in number of branchiostegal rays across taxa from above. They prompt Phenoscape to map the character changes on a phylogeny. From this they can see that all cypriniforms, including zebrafish, have three branchiostegal rays, but other fishes, including close ostariophysan relatives, have higher and lower numbers. Specifically, reduction in number has occurred multiple times; solenostomids and syngnathids (ghost pipefishes and pipefishes), giganturids, and saccopharyngoid (gulper and swallower) eels have the fewest branchiostegal rays.

Input

A list of taxa and their phenotypes for a previously searched character specification. A phylogenetic tree including the taxa of interest. [There should be trees internal to the system. And I think this should simply be combined with the previous use case--Tjvision 13:02, 15 August 2008 (EDT)]

Output

A graphical representation of the phylogenetic tree (cladogram). The branches are colored to represent reconstructed ancestral states of the given character values (or ambiguity) using, e.g. parsimony. The phenotypic data matrix is shown at the tips.

View every change in an anatomical entity mapped on a tree (ranked #2 by PM with two other similar mapping use cases, also ranked #2)

Motivation

The user may be interested in how a particular structure has evolved, without knowing what types of changes have occurred in that structure. It would be useful to view the pattern of evolution of all species phenotypes involving that structure, visualized on a phylogeny as in #Map changes of an evolutionary character on a phylogeny. [If multiple structures are examined, this could feed into a "generate me a NEXUS file" use case--Tjvision 13:20, 15 August 2008 (EDT)] [Narrowed searches (e.g. from E="Branchiostegal rays", Q="all" to Q="count" are equivalent to Use cases "Map changes of an evolutionary character on a phylogeny" and "Identify evolutionary changes that match the phenotype of a zebrafish mutant" --Pmabee@usd.edu 17:33, 18 August 2008 (EDT)]. A user may also want to broaden a search (e.g., to "E=dermal bone") in the cases where the search results are null--Pmabee@usd.edu 16:35, 18 August 2008 (EDT)]

Example

A biologist is interested in the parietal bone. She chooses this term from the anatomy ontology and then views a phylogeny displaying all reconstructed character transitions involving the parietal as an entity.

Input

An anatomical term from an anatomy ontology. A (user supplied or built-in) phylogeny containing the species of interest.

Output

A display of the phylogeny mapping changes for each phenotypic character. (Possibly a listing of all phenotypes for each species which contain the entered term as an entity).

3. Comparison of genetically and evolutionarily correlated characters

Motivation

For phylogenetic systematics, one would like to know if two or more characters show phylogenetic correlation due to an underlying genetical/developmental correlation. One way to address this is to determine if changes to both (all) those characters are found in a single-gene zebrafish mutant. [The reverse of this use case can be imagined--Tjvision 12:55, 15 August 2008 (EDT)][agree with Todd--Pmabee@usd.edu 16:03, 18 August 2008 (EDT)]

Example

The user observes a suite of changes in the size of the dentary, maxilla, ceratohyal, and opercle bones that support the monophyly of a particular clade. The user queries the system for zebrafish mutant phenotypes in which two or more of the above anatomical terms are used. Phenoscape returns sox9ahi1134, in which the the dentary, opercle, and maxilla bones are reduced in size relative to wild-type, whereas other bones are relatively unaffected. [This seems like a variant on the first use case with multiple Es and no Qs, so perhaps these can be combined --Tjvision 12:55, 15 August 2008 (EDT)]

Input

Multiple phenotype search specifications: entity, quality, etc., representing the evolutionary character changes. Option to match exactly or to match descendant terms (as in quality "count", above).[Also, option to match ancestral terms a node or two up.--Pmabee@usd.edu 16:04, 18 August 2008 (EDT)]

Output

A list of ZFIN mutant identifiers and associated genes which are associated with phenotypes matching 2 or more of the search criteria. [The output should at least be a matrix of genes against phenotypes --Tjvision 12:55, 15 August 2008 (EDT)] [Needs analysis workshop identified this "Correlation matrix of traits" as a high (#2) priority --Pmabee@usd.edu 18:10, 18 August 2008 (EDT)]

4. View values for a particular character for a set of species with some value for another character (ranked equivalently with one other use case)

Motivation

A particular phenotype change may have evolutionary consequences for another aspect of phenotype. This may be because they are linked via developmental or physical constraints, or are related through their effect on natural selection.

Example

User observes a number of species missing the parietal bone and notes that these species are also generally small in size. To determine the generality of that conclusion, the user searches for all species in the database which lack the parietal bone and requests the body length value for each. [Is this not also or alternatively done with zebrafish mutant data?--Tjvision 13:12, 15 August 2008 (EDT)]

Input

A phenotype search specification for the phenotype to match. A second phenotype search specification for the attribute for which to search for values.

Output

A table containing, in each row, the name of the taxon matching the first entered phenotype, the value of the first phenotype (included under the assumption this may vary), and the value of the second phenotype.

4. View all species with multiple phenotypes matching a condition

Motivation

To find taxa (or mutants) matching complex phenotype descriptions. [[This begins to really take advantage of ontological reasoning--Tjvision 13:16, 15 August 2008 (EDT) ]]

Example

A biologist wants to list all the species that have lost more than one bone. Or more specifically, all species that have lost more than one bone in the head.

Input

A phenotype search specification, with constraints on the entity term such as "is_a term1" and "part_of term2". The threshold number of annotations matching the phenotype required to include a taxon.

Output

A list of taxa (and/or mutants?) and their phenotypes matching the input criteria.

5. Find morphological hot spots

Motivation

Some parts of anatomy may evolve rapidly relative to others. These can be identified as those that exhibit many evolutionary changes in phenotype. This could be a simple metric like finding anatomical terms that exhibit many different value states for characters. A more complex analysis might perform ancestral state reconstruction for every character in the database, and return the entity terms involved in phenotypes with the most transitions. [Quality terms also of significant interest - is size, shape, or presence/absence the most common change in a particular clade? How does frequency of different types (size, shape, qualitative) compare across clades? compare between clades and mutants? --Pmabee@usd.edu 16:25, 18 August 2008 (EDT)]

[Are there other general properties of interest that are easily measured? Slowest evolving, most/least evolutionary data available, affected by most/least mutants, etc].

Example

The user wants to know whether structures that evolve rapidly share any genetic commonality. User obtains a list of rapidly evolving structures using this query, and then performs further analyses of those structures.

Input

All phenotypes in the database.

Output

A list of anatomy terms exhibiting the highest number of phenotypes or changes, depending on the metric.

6. Genetic or evolutionary modules: discovery and testing

Motivation

Some parts of anatomy may be developmentally integrated, sharing common genetic pathways. Structures that are identified as sharing common developmental genes and/or pathways may be grouped as a possible module; this assemblage of entities can be mapped on phylogeny to see the level of constraint on its parts (i.e. whether it is evolutionarily conserved across, e.g. more nodes than one would expect from a randomly assembled group of entities). Genetic module discovery mechanism unclear.

Example

A user may want to know whether some a priori group of entities constitute a module (testing). User may also want to know whether modules exist and what constitutes them (discovery)

Input

All phenotypes in the database.