1. SCIM: universal single-cell matching with unpaired feature sets
- Author
-
Stark, Stefan G, Ficek, Joanna, Locatello, Francesco, Bonilla, Ximena, Chevrier, Stéphane, Singer, Franziska, Aebersold, Rudolf, Al-Quaddoomi, Faisal S, Albinus, Jonas, Alborelli, Ilaria, Andani, Sonali, Attinger, Per-Olof, Bacac, Marina, Baumhoer, Daniel, Beck-Schimmer, Beatrice, Beerenwinkel, Niko, Beisel, Christian, Bernasconi, Lara, Bertolini, Anne, Bodenmiller, Bernd, Casanova, Ruben, Chicherova, Natalia, D'Costa, Maya, Danenberg, Esther, Davidson, Natalie, gan, Monica-Andreea Dră, Dummer, Reinhard, Engler, Stefanie, Erkens, Martin, Eschbach, Katja, Esposito, Cinzia, Fedier, André, Ferreira, Pedro, Frei, Anja L, Frey, Bruno, Goetze, Sandra, Grob, Linda, Gut, Gabriele, Günther, Detlef, Haberecker, Martina, Haeuptle, Pirmin, Heinzelmann-Schwarz, Viola, Herter, Sylvia, Holtackers, Rene, Huesser, Tamara, Irmisch, Anja, Jacob, Francis, Jacobs, Andrea, Jaeger, Tim M, Jahn, Katharina, James, Alva R, Jermann, Philip M, Kahles, André, Kahraman, Abdullah, Koelzer, Viktor H, Kuebler, Werner, Kuipers, Jack, Kunze, Christian P, Kurzeder, Christian, Lehmann, Kjong-Van, Levesque, Mitchell, Lugert, Sebastian, Maass, Gerd, Manz, Markus, Markolin, Philipp, Mena, Julien, Menzel, Ulrike, Metzler, Julian M, Miglino, Nicola, Milani, Emanuela S, Moch, Holger, Muenst, Simone, Murri, Riccardo, Ng, Charlotte KY, Nicolet, Stefan, Nowak, Marta, Pedrioli, Patrick GA, Pelkmans, Lucas, Piscuoglio, Salvatore, Prummer, Michael, Ritter, Mathilde, Rommel, Christian, Rosano-González, María L, Rätsch, Gunnar, Santacroce, Natascha, Castillo, Jacobo Sarabia del, Schlenker, Ramona, Schwalie, Petra C, Schwan, Severin, Schär, Tobias, Senti, Gabriela, Sivapatham, Sujana, Snijder, Berend, Sobottka, Bettina, Sreedharan, Vipin T, Stark, Stefan, Stekhoven, Daniel J, Theocharides, Alexandre PA, Thomas, Tinu M, Tolnay, Markus, Tosevski, Vinko, Toussaint, Nora C, Tuncel, Mustafa A, Tusup, Marina, Drogen, Audrey Van, Vetter, Marcus, Vlajnic, Tatjana, Weber, Sandra, Weber, Walter P, Wegmann, Rebekka, Weller, Michael, Wendt, Fabian, Wey, Norbert, Wicki, Andreas, Wollscheid, Bernd, Yu, Shuqing, Ziegler, Johanna, Zimmermann, Marc, Zoche, Martin, Zuend, Gregor, and University of Zurich
- Subjects
Statistics and Probability ,1303 Biochemistry ,AcademicSubjects/SCI01060 ,Computer science ,610 Medicine & health ,computer.software_genre ,Biochemistry ,03 medical and health sciences ,0302 clinical medicine ,Text mining ,1312 Molecular Biology ,1706 Computer Science Applications ,Humans ,Profiling (information science) ,2613 Statistics and Probability ,Molecular Biology ,030304 developmental biology ,Data ,0303 health sciences ,Sequence Analysis, RNA ,business.industry ,Gene Expression Profiling ,Autoencoder ,Computer Science Applications ,Computational Mathematics ,Computational Theory and Mathematics ,10032 Clinic for Oncology and Hematology ,Bipartite graph ,Data mining ,Single-Cell Analysis ,business ,computer ,2605 Computational Mathematics ,Algorithms ,Software ,030217 neurology & neurosurgery ,Data integration ,1703 Computational Theory and Mathematics - Abstract
Motivation: Recent technological advances have led to an increase in the production and availability of single-cell data. The ability to integrate a set of multi-technology measurements would allow the identification of biologically or clinically meaningful observations through the unification of the perspectives afforded by each technology. In most cases, however, profiling technologies consume the used cells and thus pairwise correspondences between datasets are lost. Due to the sheer size single-cell datasets can acquire, scalable algorithms that are able to universally match single-cell measurements carried out in one cell to its corresponding sibling in another technology are needed. Results: We propose Single-Cell data Integration via Matching (SCIM), a scalable approach to recover such correspondences in two or more technologies. SCIM assumes that cells share a common (low-dimensional) underlying structure and that the underlying cell distribution is approximately constant across technologies. It constructs a technology-invariant latent space using an autoencoder framework with an adversarial objective. Multi-modal datasets are integrated by pairing cells across technologies using a bipartite matching scheme that operates on the low-dimensional latent representations. We evaluate SCIM on a simulated cellular branching process and show that the cell-to-cell matches derived by SCIM reflect the same pseudotime on the simulated dataset. Moreover, we apply our method to two real-world scenarios, a melanoma tumor sample and a human bone marrow sample, where we pair cells from a scRNA dataset to their sibling cells in a CyTOF dataset achieving 90% and 78% cell-matching accuracy for each one of the samples, respectively., Bioinformatics, 36 (S2), ISSN:1367-4803, ISSN:1460-2059
- Published
- 2020