Edgar Dahl

Abstract Fast and reliable detection of patients with severe and heterogeneous illnesses is a major goal of precision medicine 1,2 . Patients with leukaemia can be identified using machine learning on the basis of their blood transcriptomes 3 . However, there is an increasing divide between what is technically possible and what is allowed, because of privacy legislation 4,5 . Here, to facilitate the integration of any medical data from any data owner worldwide without violating privacy laws, we introduce Swarm Learning—a decentralized machine-learning approach that unites edge computing, blockchain-based peer-to-peer networking and coordination while maintaining confidentiality without the need for a central coordinator, thereby going beyond federated learning. To illustrate the feasibility of using Swarm Learning to develop disease classifiers using distributed data, we chose four use cases of heterogeneous diseases (COVID-19, tuberculosis, leukaemia and lung pathologies). With more than 16,400 blood transcriptomes derived from 127 clinical studies with non-uniform distributions of cases and controls and substantial study biases, as well as more than 95,000 chest X-ray images, we show that Swarm Learning classifiers outperform those developed at individual sites. In addition, Swarm Learning completely fulfils local confidentiality regulations by design. We believe that this approach will notably accelerate the introduction of precision medicine.

Pax genes are a family of developmental control genes that encode nuclear transcription factors. They are characterized by the presence of the paired domain, a conserved amino acid motif with DNA-binding activity. Originally, paired-box-containing genes were detected in Drosophila melanogaster, where they exert multiple functions during embryogenesis. In vertebrates, Pax genes are also involved in embryogenesis. Mutations in four out of nine characterized Pax genes have been associated with either congenital human diseases such as Waardenburg syndrome (PAX3), Aniridia (PAX6), Peter's anomaly (PAX6), renal coloboma syndrome (PAX2) or spontaneous mouse mutants (undulated (Pax1), Splotch (Pax3), Small eye (Pax6), Pax2(1)Neu), which all show defects in development. Recently, analysis of spontaneous and transgenic mouse mutants has revealed that vertebrate pax genes are key regulators during organogenesis of kidney, eye, ear, nose, limb muscles, vertebral column and brain. Like their Drosophila counterparts, vertebrate Pax genes are involved in pattern formation during embryogenesis, possibly by determining the time and place of organ initiation or morphogenesis. For most tissues, however, the nature of the primary developmental action of Pax transcription factors remains to be elucidated. One predominant theme is signal transduction during tissue interactions, which may lead to a position-specific regulation of cell proliferation.

PURPOSE: CD24 is expressed in hematological malignancies as well as in a large variety of solid tumors including breast cancer. We aimed to evaluate CD24 protein expression in breast cancer and to correlate to clinicopathological data including patient survival. EXPERIMENTAL DESIGN: Primary breast carcinomas (201) with a mean clinical follow-up time of 53 months were immunostained using a monoclonal CD24 antibody (Ab-2, clone 24C02). The staining was evaluated as negative versus positive for statistical analysis. RESULTS: In invasive breast carcinomas, CD24 expression was observed in 84.6% of cases. In univariate survival analyses, a significant association of CD24 expression with shortened patient overall survival (5-year survival rate 91.9% versus 83.8%; P = 0.031; log rank test) and disease-free survival (5-year progression rate 88.3% versus 57.0%; P = 0.0008) was demonstrated. In multivariate analyses CD24, tumor grading and nodal status were significant prognostic parameters for shortened disease-free survival. CONCLUSIONS: Our data suggest that CD24 expression in primary breast cancer as detected by immunohistochemistry might be a new marker for a more aggressive breast cancer biology.