Biological Insights Knowledge Graph: an integrated knowledge graph to support drug development

Geleta, D.; Nikolov, A.; Edwards, G.; Gogleva, A.; Jackson, R.; Jansson, E.; Lamov, A.; Nilsson, S.; Pettersson, M.; Poroshin, V.; Rozemberczki, B.; Scrivener, T.; Ughetto, M.; Papa, E.

doi:10.1101/2021.10.28.466262

bioRxiv · 10.1101/2021.10.28.466262

Biological Insights Knowledge Graph: an integrated knowledge graph to support drug development

Geleta, D.·Nikolov, A.·Edwards, G.·Gogleva, A.·Jackson, R.·Jansson, E.·Lamov, A.·Nilsson, S.·Pettersson, M.·Poroshin, V.·Rozemberczki, B.·Scrivener, T.·Ughetto, M.·Papa, E.

View original source ↗Export metadata ↓

Abstract

The use of knowledge graphs as a data source for machine learning methods to solve complex problems in life sciences has rapidly become popular in recent years. Our Biological Insights Knowledge Graph (BIKG) combines relevant data for drug development from public as well as internal data sources to provide insights for a range of tasks: from identifying new targets to repurposing existing drugs. Besides the common requirements to organisational knowledge graphs such as being able to capture the domain precisely and give the users the ability to search and query the data, the focus on handling multiple use cases and supporting use case-specific machine learning models presents additional challenges: the data models must also be streamlined for the performance of downstream tasks; graph content must be easily customisable for different use cases; different projections of the graph content are required to support a wider range of different consumption modes. In this paper we describe our main design choices in implementation of the BIKG graph and discuss different aspects of its life cycle: from graph construction to exploitation.

Explore related subjects

systems biology Preprint

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Geleta, D., Nikolov, A., Edwards, G., Gogleva, A., Jackson, R., Jansson, E., Lamov, A., Nilsson, S., Pettersson, M., Poroshin, V., Rozemberczki, B., Scrivener, T., Ughetto, M., Papa, E.. 2021-11-01. Biological Insights Knowledge Graph: an integrated knowledge graph to support drug development. https://doi.org/10.1101/2021.10.28.466262

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Three-dimensional mapping of human lung fibrosis identifies spatial coupling between tissue microarchitecture and maladaptive repair states

Human lung fibrosis progressively destroys three-dimensional tissue architecture, yet how the structural microenvironment shapes cellular states remains unclear. Integrating 3D X-ray histology with spatial omics, we generate a volumetric multimodal atlas of human lung fibrosis that enables direct interrogation of relationships between tissue microarchitecture and spatial molecular programmes. We identify collapse of the alveolar-capillary unit - the coupled loss of alveolar type 1 epithelial cells, capillary endothelial cells, and basement membrane integrity - as a structural transition associated with early fibrogenesis. Surrounding transitional zones exhibit aberrant epithelial differentiation and proteolytic matrix remodelling, with epithelium shifting toward bronchiolised, airway-like states. Stromal activation converges on spatially restricted fibrogenic cores, sites of maximal collagen maturation and matrix stabilisation that function as signalling hubs within surrounding perifocal and vascular microenvironments. Molecular signatures of atlas-defined tissue states were associated with lung function and mortality across independent patient cohorts. Together, these data support a model in which fibrosis emerges as a spatial tissue-state process, with alveolar-capillary collapse redirecting repair towards mechanically stabilised niches that reinforce pathological remodelling.

systems biology↗

bioRxiv2026

A three-factor model can explain symmetry breaking and the n-1 rule during X-chromosome inactivation

X-chromosome inactivation (XCI) is a mammalian gene regulatory process that via chromosome-scale transcriptional silencing, leaves active a single X chromosome per diploid genome complement. The mechanisms that enable cells to select the appropriate number of X chromosomes to inactivate while leaving others active remain unclear. We demonstrate that all known chromosomal counting scenarios during XCI can be accounted for by a three-factor mathematical model that includes (1) Xist, the long noncoding RNA (lncRNA) responsible for gene silencing during XCI in eutherians, (2) an X-linked transcriptional activator of Xist, and (3) the autosomally-encoded SPEN, an Xist RNA-binding protein that is essential for gene silencing during XCI. In our model, symmetry between active X chromosomes is broken by stochastic differences in the number of Xist molecules that associate with the X chromosomes at the onset of XCI. Subsequently, integrated positive and negative feedback loops amplify then propagate superstoichiometric assemblies of SPEN complexed with Xist on the inactive chromosome(s), and select for levels of X-linked activator that enable only one X chromosome per diploid genome complement to remain active. Thus, asymmetric propagation of the SPEN-Xist ribonucleoprotein complex may play an essential role in symmetry breaking and chromosomal counting during XCI, independently of or in concert with cis-acting networks of DNA regulatory elements. Our results also have conceptual implications for gene regulatory circuits on autosomes.

systems biology↗

bioRxiv2026

stem.MAP: Identification of cell-type-specific protein complexes in mammalian embryonic stem cells.

Protein complexes carry out critical functions during development. Mapping cell-type-specific protein complexes in early embryogenesis is essential for uncovering the molecular mechanisms of embryonic development. We unfortunately currently lack a catalog of protein complexes specific for embryogenesis including embryonic stem cells. Here we develop a proteomics-based machine learning workflow to predict protein complexes present in mouse embryonic stem cells (mESC). Our method integrates data from a cell-type agnostic protein complex resource with co-fractionation mass spectrometry (CF-MS) experiments on mESC. Our resource, stem.MAP, identifies 3,986 complexes encompassing ~133,000 unique interactions, and 8,616 proteins. stem.MAP captures variant complexes critical for stem cell function including the PRC2, BAF and NuRD complexes. stem.MAP also discriminates mutually exclusive interactions specific to embryonic stem cells (e.g. Eif4ebp2 - Eif4e). Finally we observe 1,388 high confidence stem.MAP interactions found only in mESC CF-MS data, many with support from AlphaFold modeling and a gastruloid coregulation network. We make stem.MAP protein complexes available through our web resource: stem.proteincomplexes.org. We anticipate stem.MAP will be beneficial to the study of early development and we expect our proteomics-based machine learning workflow to be applicable to the creation of any context-specific protein interaction network.

systems biology↗