Convergence matter

[Lujw9]

Hi,
I want to perform SDA on my transcriptomic data but encountered convergence failure. I have tried to set "--num_comps" as 100, 200, 500 and 1000 with 5000 iterations, but all failed.

In my data, the gene expression of one tissue was profiled in 3 conditions: naive, treatment1 and treatment2. A total of 22411 genes were used for analysis and the standard to filter genes followed the illustration in [1].
How could I solve this problem? Please help! Thank you!!!

[1] Tensor decomposition for multiple-tissue gene expression experiments. (2016)