TY - JOUR
T1 - Transfer learning efficiently maps bone marrow cell types from mouse to human using single-cell RNA sequencing
AU - Stumpf, Patrick S.
AU - Du, Xin
AU - Imanishi, Haruka
AU - Kunisaki, Yuya
AU - Semba, Yuichiro
AU - Noble, Timothy
AU - Smith, Rosanna C.G.
AU - Rose-Zerili, Matthew
AU - West, Jonathan J.
AU - Oreffo, Richard O.C.
AU - Farrahi, Katayoun
AU - Niranjan, Mahesan
AU - Akashi, Koichi
AU - Arai, Fumio
AU - MacArthur, Ben D.
N1 - Funding Information:
This research was funded by the Medical Research Council (MC_PC_15078), the Research Management Committee at the University of Southampton, Faculty of Medicine and The Alan Turing Institute under the EPSRC grant EP/N510129/1. ROCO acknowledges support from the UK Regenerative Medicine Platform “Acellular/Smart Materials—3D Architecture” (MR/R015651/1), the Rosetrees Trust, Wessex Medical Research and the Biotechnology and Biological Sciences Research Council (BB/P017711/ 1). We would like to thank Alistair Bailey (University of Southampton) for his helpful discussion of Keras.
Publisher Copyright:
© 2020, The Author(s).
PY - 2020/12
Y1 - 2020/12
N2 - Biomedical research often involves conducting experiments on model organisms in the anticipation that the biology learnt will transfer to humans. Previous comparative studies of mouse and human tissues were limited by the use of bulk-cell material. Here we show that transfer learning—the branch of machine learning that concerns passing information from one domain to another—can be used to efficiently map bone marrow biology between species, using data obtained from single-cell RNA sequencing. We first trained a multiclass logistic regression model to recognize different cell types in mouse bone marrow achieving equivalent performance to more complex artificial neural networks. Furthermore, it was able to identify individual human bone marrow cells with 83% overall accuracy. However, some human cell types were not easily identified, indicating important differences in biology. When re-training the mouse classifier using data from human, less than 10 human cells of a given type were needed to accurately learn its representation. In some cases, human cell identities could be inferred directly from the mouse classifier via zero-shot learning. These results show how simple machine learning models can be used to reconstruct complex biology from limited data, with broad implications for biomedical research.
AB - Biomedical research often involves conducting experiments on model organisms in the anticipation that the biology learnt will transfer to humans. Previous comparative studies of mouse and human tissues were limited by the use of bulk-cell material. Here we show that transfer learning—the branch of machine learning that concerns passing information from one domain to another—can be used to efficiently map bone marrow biology between species, using data obtained from single-cell RNA sequencing. We first trained a multiclass logistic regression model to recognize different cell types in mouse bone marrow achieving equivalent performance to more complex artificial neural networks. Furthermore, it was able to identify individual human bone marrow cells with 83% overall accuracy. However, some human cell types were not easily identified, indicating important differences in biology. When re-training the mouse classifier using data from human, less than 10 human cells of a given type were needed to accurately learn its representation. In some cases, human cell identities could be inferred directly from the mouse classifier via zero-shot learning. These results show how simple machine learning models can be used to reconstruct complex biology from limited data, with broad implications for biomedical research.
UR - http://www.scopus.com/inward/record.url?scp=85097062703&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85097062703&partnerID=8YFLogxK
U2 - 10.1038/s42003-020-01463-6
DO - 10.1038/s42003-020-01463-6
M3 - Article
C2 - 33277618
AN - SCOPUS:85097062703
VL - 3
JO - Communications Biology
JF - Communications Biology
SN - 2399-3642
IS - 1
M1 - 736
ER -