Machine Learning of Bone Marrow Histopathology Identifies Genetic and Clinical Determinants in MDS Patients

1 Hematology Research Unit Helsinki, University of Helsinki and Helsinki University 8 Hospital Comprehensive Cancer Center, Helsinki, Finland 9 2 Translational Immunology Research Program, University of Helsinki, Helsinki, 10 Finland 11 3 iCAN Digital Precision Cancer Medicine Flagship, Helsinki, Finland 12 4 Helsinki University Hospital Comprehensive Cancer Center, Department of 13 Hematology, Helsinki, Finland 14 5 Institute for Molecular Medicine Finland, HiLIFE, University of Helsinki, Helsinki, 15 Finland 16 6 Department of Pathology, HUSLAB, Helsinki University Hospital and University of 17 Helsinki, Helsinki, Finland 18 7 HUS Diagnostic Center, HUSLAB, Helsinki University Hospital, Helsinki, Finland 19 8 Department of Cancer Genetics, Institute for Cancer Research, Oslo University 20 Hospital, and Oslo Centre for Biostatistics and Epidemiology, University of Oslo, 21 Oslo, Norway 22 9 Novartis Pharmaceuticals, Basel, Switzerland 23 10 Department of Clinical Chemistry and Hematology, University of Helsinki, 24 Helsinki, Finland 25 26 27 * Corresponding authors 28 † These authors contributed equally to this work. 29 30 31 Address for correspondence: 32 Drs S Mustjoki and O Brück, Hematology Research Unit Helsinki, University of 33 Helsinki and Helsinki University Hospital Comprehensive Cancer Center, 34 Haartmaninkatu 8, FIN-00290 Helsinki, Finland. E-mail: satu.mustjoki@helsinki.fi, 35 oscar.bruck@hus.fi 36 Running title: Histopathological Landscape of MDS 37 Word count: Abstract 150, Manuscript 4490 38 Number of Figures and tables: 4 figures; 1 table, 10 supplemental figures, 3 39 supplemental tables 40 41 Disclosure of Conflicts of Interest 42 K.P. received honoraria and research funding from Celgene, Incyte, Novartis and 43 Bristol-Myers Squibb. S.M. received honoraria and research funding from Pfizer, 44