MarsGT: Multi-omics analysis for rare population inference using single-cell graph transformer
Data preprocessing MarsGT initiates by inputting the raw count matrices derived from matched scRNA-seq \({{{{{{\bf{X}}}}}}}^{{{{{{\bf{R}}}}}}}=\{{x}_{{ik}}^{R}{|i}={{{{\mathrm{1,2}}}}},\ldots,{M}_{1}{;k}={{{{\mathrm{1,2}}}}},\ldots,N\}\) and scATAC-seq \({{{{{{\bf{X}}}}}}}^{{{{{{\bf{A}}}}}}}=\{{x}_{{jk}}^{A}{|j}={{{{\mathrm{1,2}}}}},\ldots,{M}_{2}{;k}={{{{\mathrm{1,2}}}}},\ldots,N\}\). For the scRNA-seq data matrix, we organize it such that rows represent genes, whereas cells constitute the columns. Conversely, the scATAC-seq data…
![Reputation-Based Federated Learning Defense to Mitigate Threats in EEG Signal Classification. (arXiv:2401.01896v1 [cs.CR])](http://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png "Reputation-Based Federated Learning Defense to Mitigate Threats in EEG Signal Classification. (arXiv:2401.01896v1 [cs.CR])"){kind=logo}
