Multiple GSEA Plots

[kirillkrs]

Hi!

I would like create GSEA Plot with multiple gene sets result.
For exemple, I have data from two analysis and I need create for each analysis GSEA Plot, e.g. green line first data, blue line second data. But I don't understand how can I do this. I know it can be realise using R, but if I can do this using Python it will be cool.

I need plot like this:

[zqfang]

This is not trivial, since the GSEA object contains everything you need to re-create figures. I can try to implement one when I have time next week

[kirillkrs]

Hi!
I tried some manipulations with code from gseaplot class of and get some result. I Just save results for needed terms in variables and then create plot with each data.

=============================

### Needed terms
terms_hdc = pre_gs_hdc.res2d.Term
data_hdc = pre_gs_hdc.results[terms_hdc[10]]
terms_lps = pre_gs_lps.res2d.Term
data_lps = pre_gs_lps.results[terms_lps[4]]

### Visualisation
fig = plt.figure(figsize=(7,5))

### GSEAplot for two subplots and it's params
rect = [0.05, 0.3, 0.8, 0.8]
ax4 = fig.add_axes(rect)
ax4.plot(np.arange(len(data_lps['RES'])), data_lps['RES'], linewidth=4, color="#009900", label='LPS')
ax4.plot(np.arange(len(data_hdc['RES'])), data_hdc['RES'], linewidth=4, color="#0033FF", label='HDC')

trans4 = transforms.blended_transform_factory(ax4.transAxes, ax4.transData)
ax4.hlines(0, 0, 1, linewidth=1, transform=trans4, color="#666666")
ax4.set_ylabel("Enrichment Score", fontsize=12, fontweight="bold")

ax4.tick_params(
    axis="both",
    which="both",
    bottom=False,
    top=False,
    right=False,
    labelbottom=False,
    labelsize=14,
)

ax4.locator_params(axis="y", nbins=10)
ax4.yaxis.set_major_formatter(
    plt.FuncFormatter(lambda tick_loc, tick_num: "{:.1f}".format(tick_loc))
)

plt.title('HALLMARK\nIL-6/JAK/STAT3 Signaling', loc='center', fontsize=16, fontweight="bold")
ax4.legend(loc="upper right", fontsize=13, facecolor="#FFFFFF")

### First of hits_subgraph
rect2_1 = [0.05, 0.25, 0.8, 0.05]
ax2_1 = fig.add_axes(rect2_1, sharex=ax1)
hit_indices2_1 = data_lps['hits']
trans2_1 = transforms.blended_transform_factory(ax2_1.transData, ax2_1.transAxes)
ax2_1.vlines(hit_indices2_1, 0, 1, linewidth=0.5, transform=trans2_1, color='#009900')
# ax2_1.vlines(hit_indices2_1, 0, 1, linewidth=0.5, transform=trans2_1, color="#FF0000")
ax2_1.spines["bottom"].set_visible(True)
ax2_1.tick_params(
    axis="y",
    which="both",
    bottom=False,
    top=False,
    right=False,
    left=False,
    labelbottom=False,
    labelleft=False,
)

### Second of hits_subgraph
rect2_2 = [0.05, 0.2, 0.8, 0.05]
ax2_2 = fig.add_axes(rect2_2, sharex=ax1)
hit_indices2_2 = data_hdc['hits']
trans2_2 = transforms.blended_transform_factory(ax2_2.transData, ax2_2.transAxes)
ax2_2.vlines(hit_indices2_2, 0, 1, linewidth=0.5, transform=trans2_2, color="#0033FF")
ax2_2.spines["bottom"].set_visible(True)
ax2_2.tick_params(
    axis="both",
    which="both",
    bottom=True,
    top=False,
    right=False,
    left=False,
    labelbottom=True,
    labelleft=False,
)

plt.show()

=============================

And result of these manipulations is plot:

---

=============================

I don't know can it be helpfull or helpless, but I wanna share this result)

[zqfang]

This is helpfull. Thank you very much. I will add this new function to the next release.