1. Introduction
Currently, about 8% of the human protein-coding genes have no function annotated in neXtProt [
1
].
For half of these ~1500 uncharacterized genes, protein products have been confidently identified [
2
].
The HUPO Human Proteome Project recently launched an initiative to understand their functions [
3
].
Whereas some biological functions can be investigated in human cell lines, others require spatial and
temporal integration of processes taking place in different cell types and can only be studied at the level
of an organism. Investigation of such complex functions is usually performed using model organisms
such as mouse or fly and the results are then extrapolated to the human protein. The choice of the model
organism is governed by scientific, technical, economic, and ethical considerations. According to the
current international regulations, organisms that do not experience pain, based on current scientific
understanding, should be used whenever it is possible. In addition, laboratories tend to favor models
that are low cost and easy to maintain. Of course, this is only possible if the protein of interest is
conserved in such organisms. This information can be found in phylogenetic databases such as OrthoDB
[
4
]. neXtProt maintains a list of ~200 tutorial SPARQL queries to support the research on human
proteins. A federated query between neXtProt and OrthoDB has been added to help researchers finding
a suitable model to characterize their proteins of interest.
2. Results and discussion
Since Drosophila melanogaster is a recognized model for neuroscience research [
5
], we built a query
that retrieves the list of uncharacterized proteins detected by immunochemistry at high levels in the
human brain that have homologs in Drosophila melanogaster. The query (NXQ_00300) has been added
to
the
list
tutorial
queries
of
the
advanced
neXtProt
SNORQL
query
tool
(https://snorql.nextprot.org/). It reuses parts of two pre-existing neXtProt tutorial queries : NXQ_00004,
that retrieves proteins detected by immunochemistry at high levels in the brain (TS-0095 in neXtProt
human anatomy vocabulary) and NXQ_00022, that retrieves entries lacking functional annotation [
6
].
The Drosophila homologs are retrieved by the OrthoDB subquery. NXQ_00300 not only retrieves the
human protein accession numbers but also the human and Drosophila gene names (Figure 1).
In a few seconds, query NXQ_00300 applied on neXtProt data release 2022-08-18 retrieves 44
human proteins. The results can be viewed as a html page or downloaded in json, xml or csv.
Users can explore Flybase [
7
] using the retrieved Drosophila gene names to find available mutants
and their phenotypes. For some of the proteins, a synonym of the Drosophila gene symbol is retrieved
from OrthoDB instead of the official gene name. For example, for KLHDC4 (NX_Q8TBB5), the
retrieved name of the Drosophila homolog is anon-WO0172774.58 instead of CG4069. This example
highlights the need of name standardization across the different resources. Fortunately, it is possible to
explore Flybase using any name or synonym for a gene. NXQ_00300 can be adapted to any other tissue
than the human brain by replacing TS-0095 by another term from the neXtProt human anatomy
vocabulary, and/or to any other model organism by replacing “Drosophila melanogaster” by the
scientific name of the organism of interest and ‘Metazoa’ by the appropriate clade.