How to explore the result?

[LJZYaaa]

Hi, very gald to have the chance to use flowsig to explore cellchat between database!
But when i run-throught the code in your tutorial, i feel confuse about the result. there are no guidelines in the tutorial to help the user explore and understand the result. feel no user-friendly :)
So could you ples give some advices to visualize the output ? @axelalmet

[axelalmet]

Hello LJZYaaa,

You're right, the plotting functionality and documentation is dreadfully lacking. I just updated flowsig to finally include some plotting functions. I plan on adding a new notebook demonstrating the different ways to plot results and updating the README.md as well by the end of this week, but in the meantime, these functions may be of use to you:

1. If you want to see which variables are differentially flowing (inflowing) or outflowing, for the pancreatic islets scRNA-seq example outlined in the README, you can run:

fig, ax = plt.subplots(figsize=(6, 4))
fs.pl.plot_differentially_flowing_signals(adata,
                                        condition_key = 'Condition',
                                        pert_key = 'IFNg',
                                        var_type = 'outflow',
                                        flowsig_expr_key = 'X_flow_orig',
                                        flowsig_network_key = 'flowsig_network_orig',
                                        qval_threshold = 0.05,
                                        logfc_threshold = 0.5,
                                        label_lowqval = True,
                                        ax=ax
                                        )
plt.show()

Setting pert_key='IFNg orients results so that those upregulated in the perturbed condition have positive log fold change. Note also that you also need to set flowsig_expr_key = 'X_flow_orig' and flowsig_network_key = 'flowsig_network_orig' to visualise both significant and non-significant variables.

2. To plot the global intercellular flow network, you can run something like this:

flow_network = fs.tl.construct_intercellular_flow_network(adata,
                                                            flowsig_network_key = 'flowsig_network',
                                                            adjacency_key = 'adjacency_validated_filtered')

fig, ax = plt.subplots(figsize=(6, 4))
fs.pl.plot_intercellular_flows(adata,
                                flow_network = flow_network,
                                flowsig_network_key = 'flowsig_network',
                                align_mode = 'horizontal',
                                width_scale = 2.0,
                                x_margin_offset = 0.3,
                                y_margin_offset = 0.0,
                                ax=ax)
plt.show()

This command plots the entire intercellular flow network. If you want to look at the flows induced by a specific inflow variable, you can set inflow_vars, e.g.,

fig, ax = plt.subplots(figsize=(6, 4))
fs.pl.plot_intercellular_flows(adata,
                                flow_network = flow_network,
                                inflow_vars = ['SSTR2'], 
                                flowsig_network_key = 'flowsig_network',
                                align_mode = 'horizontal',
                                width_scale = 2.0,
                                x_margin_offset = 0.3,
                                y_margin_offset = 0.0,
                                ax=ax)
plt.show()

You can also specify module_vars and/or outflow_vars.

More details will come ASAP!

[LJZYaaa]

Glad to recevice your reply, and can't wait to use the new version of flowsig!
Hope anything goes well on you ^^

[wolfQK]

Hi axelalmet，
Could you please let me know when you might be able to update the code for the visualization part? I really appreciate your work and look forward to the update.
Thank you very much!
@axelalmet

[Wenbin1994]

Hi! Thank you for making this tool. Excuse me, how do I reproduce the figures from the article, such as Fig3c, Fig4a, Fig5a and b, and how do I take the transcription factors from the GEM module and make figures similar to Fig6d, e and f?

[XiaolongYang-HZAU]

Thank you for developing such an excellent software! I also hope to replicate the visualization of Fig6 in your article. Is there currently visualization provided in the package, or would you be able to update the visualization section in the tutorial? Thank you very much!