Regression plane concept for analysing continuous cellular processes with machine learning

Abel Szkalisity; Filippo Piccinini; Attila Beleon; Tamas Balassa; Istvan Gergely Varga; Ede Migh; Csaba Molnar; Lassi Paavolainen; Sanna Timonen; Indranil Banerjee; Elina Ikonen; Yohei Yamauchi; Istvan Ando; Jaakko Peltonen; Vilja Pietiäinen; Viktor Honti; Peter Horvath

doi:10.1038/s41467-021-22866-x

Regression plane concept for analysing continuous cellular processes with machine learning

Abel Szkalisity
, Filippo Piccinini
, Attila Beleon
, Tamas Balassa
, Istvan Gergely Varga
, Ede Migh
, Csaba Molnar
, Lassi Paavolainen
, Sanna Timonen
, Indranil Banerjee
, Elina Ikonen
, Yohei Yamauchi
, Istvan Ando
, Jaakko Peltonen
, Vilja Pietiäinen
, Viktor Honti
, Peter Horvath^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › Scientific › peer-review

12 Citations (Scopus)

146 Downloads (Pure)

Abstract

Biological processes are inherently continuous, and the chance of phenotypic discovery is significantly restricted by discretising them. Using multi-parametric active regression we introduce the Regression Plane (RP), a user-friendly discovery tool enabling class-free phenotypic supervised machine learning, to describe and explore biological data in a continuous manner. First, we compare traditional classification with regression in a simulated experimental setup. Second, we use our framework to identify genes involved in regulating triglyceride levels in human cells. Subsequently, we analyse a time-lapse dataset on mitosis to demonstrate that the proposed methodology is capable of modelling complex processes at infinite resolution. Finally, we show that hemocyte differentiation in Drosophila melanogaster has continuous characteristics.

Original language	English
Article number	2532
Number of pages	9
Journal	Nature Communications
Volume	12
Issue number	1
DOIs	https://doi.org/10.1038/s41467-021-22866-x
Publication status	Published - Dec 2021
MoE publication type	A1 Journal article-refereed

Funding

The authors thank Antti Lehmussola and Pekka Ruusuvuori (Tampere University of Technology, Finland) for the information provided about the SIMCEP software; Samuli Ripatti and Ida Surakka (FIMM, University of Helsinki, Finland) for their valuable comments on our experiments related to the genes associated with dyslipidemia; Anna Uro (Faculty of Medicine, University of Helsinki) for providing expertise in the biochemical quantification of lipid levels; Mariliina Arjama (FIMM, University of Helsinki, Finland) for technical expertise in cell culture; the FIMM High Throughput Biomedicine Unit for providing access to high throughput robotics and siRNA library and the FIMM High Content Imaging and Analysis unit for HC-imaging (HiLIFE, University of Helsinki and Biocenter Finland); Olli Kallioniemi (FIMM, University of Helsinki, Finland) for support in HC-imaging capabilities and funding; Gabriella Tick and Máté Görbe (BRC, Szeged, Hungary) for their help with the software documentation; the Finnish Grid and Cloud Infrastructure (urn:nbn:fi:research-infras-2016072533) for computational resources; Dóra Bokor (BRC, Szeged, Hungary) for proofreading the manuscript. A.Sz., B.T., A.B., E.M., Cs.M. and P.H. acknowledge support from the Hungarian National Brain Research Program (MTA-SE-NAP B-BIOMAG), from the LENDULET-BIOMAG Grant (2018-342), from the European Regional Development Funds (GINOP-2.3.2-15-2016-00006, GINOP-2.3.2-15-2016-00026, GINOP-2.3.2-15-2016-00037), from the H2020 (ERAPERMED-COMPASS, DiscovAIR) and from the Chan Zuckerberg Initiative (Deep Visual Proteomics). A.Sz. and E.I. acknowledges support from University of Helsinki (Centre of Excellence matching funds) and Academy of Finland (project 324929). V.P., L.P. and P.H. acknowledge support from the Finnish TEKES FiDiPro Fellow Grant 40294/13 and FIMM High Content Imaging and Analysis Unit (FIMM-HCA; HiLIFE-HELMI) and Biocenter Finland, Finnish Cancer Society, Juselius Foundation, Academy of Finland Centre of Excellence in Translational Cancer Biology, Kymenlaakso and Finnish Cultural Foundation. V.P. acknowledges University of Helsinki post-doctoral research project grant. F.P. acknowledges support from the Union for International Cancer Control (UICC) for a UICC Yamagiwa-Yoshida (YY) Memorial International Cancer Study Grant (ref: UICC-YY/678329). V.H. and I.A. acknowledge the Hungarian National Research Fund (OTKA NKFI‐2 NN118207). V.H. acknowledges support from the National Research, Development and Innovation Office (OTKA K-131484). L.P. and J.P. acknowledge support from the Academy of Finland, decision numbers 295694, 313748, 327352 and 310552.

Access to Document

10.1038/s41467-021-22866-x

Regression plane concept.s41467-021-22866-xFinal published version, 4.52 MBLicence: CC BY

Cite this

Szkalisity, A., Piccinini, F., Beleon, A., Balassa, T., Varga, I. G., Migh, E., Molnar, C., Paavolainen, L., Timonen, S., Banerjee, I., Ikonen, E., Yamauchi, Y., Ando, I., Peltonen, J., Pietiäinen, V., Honti, V., & Horvath, P. (2021). Regression plane concept for analysing continuous cellular processes with machine learning. Nature Communications, 12(1), Article 2532. https://doi.org/10.1038/s41467-021-22866-x

@article{3f52036d3a0e4536b10189182b4f02a9,

title = "Regression plane concept for analysing continuous cellular processes with machine learning",

abstract = "Biological processes are inherently continuous, and the chance of phenotypic discovery is significantly restricted by discretising them. Using multi-parametric active regression we introduce the Regression Plane (RP), a user-friendly discovery tool enabling class-free phenotypic supervised machine learning, to describe and explore biological data in a continuous manner. First, we compare traditional classification with regression in a simulated experimental setup. Second, we use our framework to identify genes involved in regulating triglyceride levels in human cells. Subsequently, we analyse a time-lapse dataset on mitosis to demonstrate that the proposed methodology is capable of modelling complex processes at infinite resolution. Finally, we show that hemocyte differentiation in Drosophila melanogaster has continuous characteristics.",

author = "Abel Szkalisity and Filippo Piccinini and Attila Beleon and Tamas Balassa and Varga, \{Istvan Gergely\} and Ede Migh and Csaba Molnar and Lassi Paavolainen and Sanna Timonen and Indranil Banerjee and Elina Ikonen and Yohei Yamauchi and Istvan Ando and Jaakko Peltonen and Vilja Pieti{\"a}inen and Viktor Honti and Peter Horvath",

note = "Funding Information: The authors thank Antti Lehmussola and Pekka Ruusuvuori (Tampere University of Technology, Finland) for the information provided about the SIMCEP software; Samuli Ripatti and Ida Surakka (FIMM, University of Helsinki, Finland) for their valuable comments on our experiments related to the genes associated with dyslipidemia; Anna Uro (Faculty of Medicine, University of Helsinki) for providing expertise in the biochemical quantification of lipid levels; Mariliina Arjama (FIMM, University of Helsinki, Finland) for technical expertise in cell culture; the FIMM High Throughput Biomedicine Unit for providing access to high throughput robotics and siRNA library and the FIMM High Content Imaging and Analysis unit for HC-imaging (HiLIFE, University of Helsinki and Biocenter Finland); Olli Kallioniemi (FIMM, University of Helsinki, Finland) for support in HC-imaging capabilities and funding; Gabriella Tick and M{\'a}t{\'e} G{\"o}rbe (BRC, Szeged, Hungary) for their help with the software documentation; the Finnish Grid and Cloud Infrastructure (urn:nbn:fi:research-infras-2016072533) for computational resources; D{\'o}ra Bokor (BRC, Szeged, Hungary) for proofreading the manuscript. A.Sz., B.T., A.B., E.M., Cs.M. and P.H. acknowledge support from the Hungarian National Brain Research Program (MTA-SE-NAP B-BIOMAG), from the LENDULET-BIOMAG Grant (2018-342), from the European Regional Development Funds (GINOP-2.3.2-15-2016-00006, GINOP-2.3.2-15-2016-00026, GINOP-2.3.2-15-2016-00037), from the H2020 (ERAPERMED-COMPASS, DiscovAIR) and from the Chan Zuckerberg Initiative (Deep Visual Proteomics). A.Sz. and E.I. acknowledges support from University of Helsinki (Centre of Excellence matching funds) and Academy of Finland (project 324929). V.P., L.P. and P.H. acknowledge support from the Finnish TEKES FiDiPro Fellow Grant 40294/13 and FIMM High Content Imaging and Analysis Unit (FIMM-HCA; HiLIFE-HELMI) and Biocenter Finland, Finnish Cancer Society, Juselius Foundation, Academy of Finland Centre of Excellence in Translational Cancer Biology, Kymenlaakso and Finnish Cultural Foundation. V.P. acknowledges University of Helsinki post-doctoral research project grant. F.P. acknowledges support from the Union for International Cancer Control (UICC) for a UICC Yamagiwa-Yoshida (YY) Memorial International Cancer Study Grant (ref: UICC-YY/678329). V.H. and I.A. acknowledge the Hungarian National Research Fund (OTKA NKFI‐2 NN118207). V.H. acknowledges support from the National Research, Development and Innovation Office (OTKA K-131484). L.P. and J.P. acknowledge support from the Academy of Finland, decision numbers 295694, 313748, 327352 and 310552. Publisher Copyright: {\textcopyright} 2021, The Author(s). Copyright: Copyright 2021 Elsevier B.V., All rights reserved.",

year = "2021",

month = dec,

doi = "10.1038/s41467-021-22866-x",

language = "English",

volume = "12",

journal = "Nature Communications",

issn = "2041-1723",

publisher = "Nature Publishing Group",

number = "1",

}

Szkalisity, A, Piccinini, F, Beleon, A, Balassa, T, Varga, IG, Migh, E, Molnar, C, Paavolainen, L, Timonen, S, Banerjee, I, Ikonen, E, Yamauchi, Y, Ando, I, Peltonen, J, Pietiäinen, V, Honti, V & Horvath, P 2021, 'Regression plane concept for analysing continuous cellular processes with machine learning', Nature Communications, vol. 12, no. 1, 2532. https://doi.org/10.1038/s41467-021-22866-x

TY - JOUR

T1 - Regression plane concept for analysing continuous cellular processes with machine learning

AU - Szkalisity, Abel

AU - Piccinini, Filippo

AU - Beleon, Attila

AU - Balassa, Tamas

AU - Varga, Istvan Gergely

AU - Migh, Ede

AU - Molnar, Csaba

AU - Paavolainen, Lassi

AU - Timonen, Sanna

AU - Banerjee, Indranil

AU - Ikonen, Elina

AU - Yamauchi, Yohei

AU - Ando, Istvan

AU - Peltonen, Jaakko

AU - Pietiäinen, Vilja

AU - Honti, Viktor

AU - Horvath, Peter

N1 - Funding Information: The authors thank Antti Lehmussola and Pekka Ruusuvuori (Tampere University of Technology, Finland) for the information provided about the SIMCEP software; Samuli Ripatti and Ida Surakka (FIMM, University of Helsinki, Finland) for their valuable comments on our experiments related to the genes associated with dyslipidemia; Anna Uro (Faculty of Medicine, University of Helsinki) for providing expertise in the biochemical quantification of lipid levels; Mariliina Arjama (FIMM, University of Helsinki, Finland) for technical expertise in cell culture; the FIMM High Throughput Biomedicine Unit for providing access to high throughput robotics and siRNA library and the FIMM High Content Imaging and Analysis unit for HC-imaging (HiLIFE, University of Helsinki and Biocenter Finland); Olli Kallioniemi (FIMM, University of Helsinki, Finland) for support in HC-imaging capabilities and funding; Gabriella Tick and Máté Görbe (BRC, Szeged, Hungary) for their help with the software documentation; the Finnish Grid and Cloud Infrastructure (urn:nbn:fi:research-infras-2016072533) for computational resources; Dóra Bokor (BRC, Szeged, Hungary) for proofreading the manuscript. A.Sz., B.T., A.B., E.M., Cs.M. and P.H. acknowledge support from the Hungarian National Brain Research Program (MTA-SE-NAP B-BIOMAG), from the LENDULET-BIOMAG Grant (2018-342), from the European Regional Development Funds (GINOP-2.3.2-15-2016-00006, GINOP-2.3.2-15-2016-00026, GINOP-2.3.2-15-2016-00037), from the H2020 (ERAPERMED-COMPASS, DiscovAIR) and from the Chan Zuckerberg Initiative (Deep Visual Proteomics). A.Sz. and E.I. acknowledges support from University of Helsinki (Centre of Excellence matching funds) and Academy of Finland (project 324929). V.P., L.P. and P.H. acknowledge support from the Finnish TEKES FiDiPro Fellow Grant 40294/13 and FIMM High Content Imaging and Analysis Unit (FIMM-HCA; HiLIFE-HELMI) and Biocenter Finland, Finnish Cancer Society, Juselius Foundation, Academy of Finland Centre of Excellence in Translational Cancer Biology, Kymenlaakso and Finnish Cultural Foundation. V.P. acknowledges University of Helsinki post-doctoral research project grant. F.P. acknowledges support from the Union for International Cancer Control (UICC) for a UICC Yamagiwa-Yoshida (YY) Memorial International Cancer Study Grant (ref: UICC-YY/678329). V.H. and I.A. acknowledge the Hungarian National Research Fund (OTKA NKFI‐2 NN118207). V.H. acknowledges support from the National Research, Development and Innovation Office (OTKA K-131484). L.P. and J.P. acknowledge support from the Academy of Finland, decision numbers 295694, 313748, 327352 and 310552. Publisher Copyright: © 2021, The Author(s). Copyright: Copyright 2021 Elsevier B.V., All rights reserved.

PY - 2021/12

Y1 - 2021/12

N2 - Biological processes are inherently continuous, and the chance of phenotypic discovery is significantly restricted by discretising them. Using multi-parametric active regression we introduce the Regression Plane (RP), a user-friendly discovery tool enabling class-free phenotypic supervised machine learning, to describe and explore biological data in a continuous manner. First, we compare traditional classification with regression in a simulated experimental setup. Second, we use our framework to identify genes involved in regulating triglyceride levels in human cells. Subsequently, we analyse a time-lapse dataset on mitosis to demonstrate that the proposed methodology is capable of modelling complex processes at infinite resolution. Finally, we show that hemocyte differentiation in Drosophila melanogaster has continuous characteristics.

AB - Biological processes are inherently continuous, and the chance of phenotypic discovery is significantly restricted by discretising them. Using multi-parametric active regression we introduce the Regression Plane (RP), a user-friendly discovery tool enabling class-free phenotypic supervised machine learning, to describe and explore biological data in a continuous manner. First, we compare traditional classification with regression in a simulated experimental setup. Second, we use our framework to identify genes involved in regulating triglyceride levels in human cells. Subsequently, we analyse a time-lapse dataset on mitosis to demonstrate that the proposed methodology is capable of modelling complex processes at infinite resolution. Finally, we show that hemocyte differentiation in Drosophila melanogaster has continuous characteristics.

UR - https://www.scopus.com/pages/publications/85105357600

U2 - 10.1038/s41467-021-22866-x

DO - 10.1038/s41467-021-22866-x

M3 - Article

C2 - 33953203

AN - SCOPUS:85105357600

SN - 2041-1723

VL - 12

JO - Nature Communications

JF - Nature Communications

IS - 1

M1 - 2532

ER -

Regression plane concept for analysing continuous cellular processes with machine learning

Abstract

Funding

Access to Document

Other files and links

Fingerprint

Cite this