Rerun recount2 notebooks

[ajlee21]

This PR is rerunning the notebooks using all SRA data from recount2. No code was changed in this PR. This PR is re-running the notebooks to get the cell outputs that were previously cleared. This should address #25

Minor changes include:

• Rename the analysis direction from human_analysis to human_general_analysis to make way for adding a new analysis where we're training only on cancer-specific gene expression data. As opposed to recount2, which contains general gene expression data.

Note:

• I tried to train the model on 20 epochs as opposed to 40, since 20 seems sufficient based on the log plot. When I was running the notebook the internet connection crashed so the model training progress was lost. I have made an issue here. Since models were already trained and uploaded in this PR, I used these files to run the analysis notebook.

[dongbohu]

Looks good!