Federated Learning With Only Positive Labels by Exploring Label Correlations.

Federated learning (FL) aims to collaboratively learn a model by using the data from multiple users under privacy constraints. In this article, we study the multilabel classification (MLC) problem under the FL setting, where trivial solution and extremely poor performance may be obtained, especially when only positive data with respect to a single class label is provided for each client. This issue can be addressed by adding a specially designed regularizer on the server side.

Inhibition of SMAD3 effectively reduces ADAMTS-5 expression in the early stages of osteoarthritis.

Objective: As one of the most important protein-degrading enzymes, ADAMTS-5 plays an important role in the regulation of cartilage homeostasis, while miRNA-140 is specifically expressed in cartilage, which can inhibit the expression of ADAMTS-5 and delay the progression of OA (osteoarthritis). SMAD3 is a key protein in the TGF-β signaling pathway, inhibiting the expression of miRNA-140 at the transcriptional and post-transcriptional levels, and studies have confirmed the high expression of SMAD3 in knee cartilage degeneration, but whether SMAD3 can mediate the expression of miRNA-140 to regulate ADAMTS-5 remains unknown.

Methods: Sprague-Dawley (SD) rat chondrocytes were extracted in vitro and treated with a SMAD3 inhibitor (SIS3) and miRNA-140 mimics after IL-1 induction.