Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

U-BIOPRED accessible handprint: Combining omics platforms to identify stable asthma subphenotypes

 
: Meulder, Bertrand de; Yen, Romain Tching Chi; Li, Chuan-Xing; Wheelock, Asa; Chung, Kian Fan; Adcock, Ian M.; Djukanović, Ratko; Wagers, Scott; Riley, John; Erpenbeck, Veit J.; Bakke, Per; Caruso, Massimo; Chanez, Pascal; Dahlén, Sven-Erik; Fowler, Stephen J.; Horváth, Ildikó; Krug, Norbert; Montuschi, Paolo; Sanak, Marek; Sandström, Thomas; Shaw, Dominic; Singer, Florian; Pandis, Ioannis; Bansal, Aruna T.; Sterk, Peter J.; Baribaud, Frédéric; Auffray, Charles

:

European Respiratory Journal 52 (2018), Supplement 62, Abstract OA3578
ISSN: 0903-1936
ISSN: 1399-3003
European Respiratory Society (ERS International Congress) <2018, Paris>
Englisch
Abstract
Fraunhofer ITEM ()

Abstract
Background: The U-BIOPRED hypothesis is that performing clustering based on multiple omics platforms (handprints) is relevant to address the unmet need in severe asthma classification.
Methods: We gathered the blood compartment-related (blood and urine) data in our adult asthma cohort (Shaw, ERJ 2015). We used Similarity Network Fusion and stability assessment to identify optimal omics platforms and cluster numbers combinations. We selected based on deviation from ideal stability (DIS).
Results: A combination of transcriptomics, proteomics, lipidomics and metabolomics is giving the best results for K= 8 and 17 (DIS = 0.06 and 0.04). The comparison of patient’s allocation between the two shows separation in K=17 of clusters in K=8 (see figure). The comparison of clinical variables shows expected differences (immune cell numbers, BMI, FEV1) but other variables are different (medication, comorbidities, biomarkers) and some clusters identified don’t show any extreme values.
Conclusion and perspectives: We have identified stable clusters of asthma patients within our cohort by multi-omics integration. The identification of molecular signatures and production of a predictive model is under way. The next steps are the validation with longitudinal measurements and external cohorts of severe asthma.

: http://publica.fraunhofer.de/dokumente/N-559363.html