UX settings of filtration properties

Filtering properties are very important objects of the system, so the handling of this objects in the Front-End UX logic requires wide spectrum of settings. They form essential part of Property Status structure, their congiguration is discussed in Configuration of filtration schema API.

The settings are:

  • title, optional string - human readable description of unit, can be not defined if name is sufficient

  • render_mode - optional string; used in UI to represent values of unit, see details below

  • tooltip - optional string; explanation of unit meaning to render in UI as tooltip (title in HTML terms)

  • classes - list of integers

    Integers in the list are index values in descriptor structurr unit-classes from Dataset Descriptor structure.

    The last setting requires an extended explanation.

Properties classification

Since there are too many various filtering properties in the system, we make an attempt to classify properties by their "universal meanings". The classification is not so clear and distinct, however it is possible to use it to separate properties for use them in different contexts.

The classification has two layers:

  • domain (facet) layer: there are numerous domains of classification, each property needs to be classified in each domain

  • low layer: in context of each domain there is list of of types

The classification applied to the dataset forms unit-classes field of Dataset Descriptor structure that is return value of dsinfo.

The facet indexes of concrete filtering property form classes field of Property Status structure.

In terms of Variable registry facet classification is defined in terms of parameters facet1, facet2, facet3 - optional strings.

The Front End can use this classification to help the user to keep unneccessary properties hidden in the filtration process.

The current version of the system provides the following classification:

facet

Title

Technical name

Knowledge Domain

facet1

Call Annotations

call

Phenotypic Data

phenotype

Compound Rules

rules

Population Genetics

popgen

Functional

function

Animal Genetics

animal

Human Genetics

human

Epigenetics

epigenetics

N/A

na1

Scale

facet2

Transcript

transcript

Variant

variant

Position

position

Window

window

Gene

gene

N/A

na2

Method

facet3

Statistical Genetics Evidence

statgen

Bioinformatics Inference

bioinf

Experimental, in Vivo

in-vivo

Experimental, in Vitro

in-vitro

Experimental, Other

explanation

Clinical Evidence

clinical

Raw Data

raw

N/A

na3

Render modes

  • Render modes for numeric properties

    • " scale, bounds "

      scale: linear/log - linear or logariphmic scale

      bounds: </>/= - meaningful bounds:

      < only upper/ > only lower / = both

      Examples:

      • "linear,<" linear scale, only upper bound is meaningful

      • "log,>" logariphmic scale, only lower bound is meaningful

    • "neighborhood" needs center/radius to determine diapason (linear scale assumed)

  • Render modes for enumerated properties

See also

Property Status structure

dsinfo

Configuration of filtration schema API