Publication:
Unsupervised machine learning identifies biomarkers of disease progression in post-kala-azar dermal leishmaniasis in Sudan

dc.contributor.authorTorres Garcia, Ana Maria
dc.contributor.authorYounis, Brima Musa
dc.contributor.authorTesema, Samuel
dc.contributor.authorSolana, Jose Carlos
dc.contributor.authorMoreno, Javier
dc.contributor.authorMartin-Galiano, Antonio Javier
dc.contributor.authorMusa, Ahmed Mudawi
dc.contributor.authorAlves, Fabiana
dc.contributor.authorCarrillo, Eugenia
dc.contributor.funderInstituto de Salud Carlos III
dc.contributor.funderCentro de Investigación Biomédica en Red - CIBERINFEC (Enfermedades Infecciosas)
dc.date.accessioned2025-05-13T09:50:59Z
dc.date.available2025-05-13T09:50:59Z
dc.date.issued2025-03
dc.description.abstractBackground: Post-kala-azar dermal leishmaniasis (PKDL) appears as a rash in some individuals who have recovered from visceral leishmaniasis caused by Leishmania donovani. Today, basic knowledge of this neglected disease and how to predict its progression remain largely unknown. Methods and findings: This study addresses the use of several biochemical, haematological and immunological variables, independently or through unsupervised machine learning (ML), to predict PKDL progression risk. In 110 patients from Sudan, 31 such factors were assessed in relation to PKDL disease state at the time of diagnosis: progressive (worsening) versus stable. To identify key factors associated with PKDL worsening, we used both a conventional statistical approach and multivariate analysis through unsupervised ML. The independent use of these variables had limited power to predict skin lesion severity in a baseline examination. In contrast, the unsupervised ML approach identified a set of 10 non-redundant variables that was linked to a 3.1 times higher risk of developing progressive PKDL. Three of these clustering factors (low albumin level, low haematocrit and low IFN-γ production in PBMCs after Leishmania antigen stimulation) were remarkable in patients with progressive disease. Dimensionality re-establishment identified 11 further significantly modified factors that are also important to understand the worsening phenotype. Our results indicate that the combination of anaemia and a weak Th1 immunological response is likely the main physiological mechanism that leads to progressive PKDL. Conclusions: A combination of 14 biochemical variables identified by unsupervised ML was able to detect a worsening PKDL state in Sudanese patients. This approach could prove instrumental to train future supervised algorithms based on larger patient cohorts both for a more precise diagnosis and to gain insight into fundamental aspects of this complication of visceral leishmaniasis.
dc.description.peerreviewed
dc.description.sponsorshipThis work was funded by DNDi via agreement MVP322/19 with ISCIII (to E.C.) and by the Instituto de Salud Carlos III through the ISCIII-AES project (PI22/00009 to E.C.). J.C. was supported by a contract awarded by CIBERINFEC (CB21/13/00018). The funders had no role in the study design, data collection and analysis, decision to publish, or prepara tion of the manuscript.
dc.format.number3
dc.format.pagee0012924
dc.format.volume19
dc.identifier.citationTorres A, Younis BM, Tesema S, Solana JC, Moreno J, Martín-Galiano AJ, Musa AM, Alves F, Carrillo E. Unsupervised machine learning identifies biomarkers of disease progression in post-kala-azar dermal leishmaniasis in Sudan. PLoS Negl Trop Dis. 2025 Mar 11;19(3):e0012924.
dc.identifier.doi10.1371/journal.pntd.0012924
dc.identifier.e-issn1935-2735
dc.identifier.issn1935-2727
dc.identifier.journalPLoS neglected tropical diseases
dc.identifier.pubmedID40067811
dc.identifier.urihttps://hdl.handle.net/20.500.12105/26655
dc.language.isoeng
dc.publisherPublic Library of Science (PLOS)
dc.relation.projectIDinfo:eu-repo/grantAgreement/ESMVP322/19
dc.relation.projectIDinfo:eu-repo/grantAgreement/ESPI22/00009
dc.relation.projectIDinfo:eu-repo/grantAgreement/ESCB21/13/00018
dc.relation.publisherversionhttps://doi.org/10.1371/journal.pntd.0012924
dc.repisalud.centroISCIII::Centro Nacional de Microbiología (CNM)
dc.repisalud.centroISCIII::Unidades Centrales Científico-Técnicas (UCCTs)
dc.repisalud.institucionISCIII
dc.rights.accessRightsopen access
dc.rights.licenseAttribution 4.0 International
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/
dc.subject.meshAdolescent
dc.subject.meshAdult
dc.subject.meshBiomarkers
dc.subject.meshChild
dc.subject.meshDisease Progression
dc.subject.meshFemale
dc.subject.meshHumans
dc.subject.meshLeishmania donovani
dc.subject.meshLeishmaniasis, Cutaneous
dc.subject.meshLeishmaniasis, Visceral
dc.subject.meshMachine Learning
dc.subject.meshMale
dc.subject.meshMiddle Aged
dc.subject.meshSudan
dc.subject.meshYoung Adult
dc.titleUnsupervised machine learning identifies biomarkers of disease progression in post-kala-azar dermal leishmaniasis in Sudan
dc.typeresearch article
dc.type.hasVersionVoR
dspace.entity.typePublication
relation.isAuthorOfPublication06e584a4-3fba-4a65-9e48-855db4852f3d
relation.isAuthorOfPublication2e690bd6-6f63-4aa1-8136-7f9b7da93062
relation.isAuthorOfPublication831dbac1-fcb6-444a-90e1-4b562eecb934
relation.isAuthorOfPublicationa05be95c-83ed-4217-ab14-cbca92cd5279
relation.isAuthorOfPublicationbb8f1dfe-8a72-4881-a19f-0b83b988f1bb
relation.isAuthorOfPublication.latestForDiscovery06e584a4-3fba-4a65-9e48-855db4852f3d
relation.isFunderOfPublication7d739953-4b68-4675-b5bb-387a9ab74b66
relation.isFunderOfPublication0b5a6d3a-4d4d-4346-9bcc-9359bbe13e72
relation.isFunderOfPublication.latestForDiscovery7d739953-4b68-4675-b5bb-387a9ab74b66
relation.isPublisherOfPublicationa2759e3d-0d58-4e8a-9fcd-c6130ee333d1
relation.isPublisherOfPublication.latestForDiscoverya2759e3d-0d58-4e8a-9fcd-c6130ee333d1

Files

Original bundle

Now showing 1 - 2 of 2
Loading...
Thumbnail Image
Name:
UnsupervisedMachineLearningIdentifiesBiomarkers_2025.pdf
Size:
1.26 MB
Format:
Adobe Portable Document Format
Loading...
Thumbnail Image
Name:
Suplementary_UnsupervisedMachineLearningIdentifiesBiomarkers_2025.zip
Size:
554.05 KB
Format: