Alternatively, the factors can also be obtained from external sc- and sn-RNA-seq references datasets, or from other spatially agnostic methods such as Seurat16 and Scanpy24 that learn factors based on grids. FICTURE can use these external factors as priors to initialize the LDA factors or use them directly as input factors.
How could I use prior knowledge to initialize the LDA factors?
Hi,
Thank you for this wonderful tool.
In the paper you write:
How could I use prior knowledge to initialize the LDA factors?
Thank you very much!