Troubleshooting x and y dimension error

[gprahalad]

Hello, thank you for developing SCARlink! I am keen on applying it to some the data I have. In line with prescribed pipeline, I ran the following commands but have had a few hiccups along the way and have hit an error I have been unsuccessful in troubleshooting, I am wondering if you perhaps have an idea of where I should be looking to resolve this.

scarlink_processing --scrna scrna_adata.h5ad --scatac scatac_adata.h5ad -o multiome_out

Using pre-computed GEX h5ad and ArchR peaks h5ad, matched cells and filtered. Cast Int64 to float due to newer h5ad version.

scarlink -o multiome_out -g hg38 scarlink -o multiome_out -g hg38 -c anno_nuc_v1_gex

No cell_info.txt was produced here (not sure if this is expected).

scarlink_tiles -o multiome_out -c anno_nuc_v1_gex

No errors produced here.

scv.plot_scarlink_output(scarlink_out, genes=['EMP1','NFKB1'] celltype='anno_nuc_v1_gex')

I receive the following dimension mismatch error:

Traceback (most recent call last): File "/software/miniforge3/envs/scarlink-env/bin/scarlink_plot", line 33, in <module> sys.exit(load_entry_point('scarlink', 'console_scripts', 'scarlink_plot')()) File "/software/SCARlink/scarlink/run_scarlink_visualization.py", line 21, in main scv.plot_scarlink_output(scarlink_out, genes=genes, File "/software/SCARlink/scarlink/src/visualization.py", line 37, in plot_scarlink_output rm.plot_gene(gene, groups=celltype, plot_frags=plot_frags, to_save=to_save, plot_dir=scarlink_out['plot_dir'], cmap=cmap, save_format=save_format, figsize=figsize, sort_gex=sort_gex, show_yticks=show_yticks, plot_shap=plot_shap, shap_cmap=shap_cmap, cluster_order=cluster_order, bg_transparent=bg_transparent) File "/software/SCARlink/scarlink/src/model.py", line 862, in plot_gene ax[c][0].plot(plotrange, X_c, color = cmap[clust]) File "/software/miniforge3/envs/scarlink-env/lib/python3.8/site-packages/matplotlib/axes/_axes.py", line 1688, in plot lines = [*self._get_lines(*args, data=data, **kwargs)] File "/software/miniforge3/envs/scarlink-env/lib/python3.8/site-packages/matplotlib/axes/_base.py", line 311, in __call__ yield from self._plot_args( File "/software/miniforge3/envs/scarlink-env/lib/python3.8/site-packages/matplotlib/axes/_base.py", line 504, in _plot_args raise ValueError(f"x and y must have same first dimension, but " ValueError: x and y must have same first dimension, but have shapes (1072,) and (232, 1)

I am quite unfamiliar with this data structure so apologies. As I vary the genes the second number of the mismatch varies but as there is already a sparsity filter I imagine this is not the sole source of the error. Any advice would be extremely helpful, thank you very much!

[gprahalad]

Resolved by forcing conformation between shapes - output seems reasonable though I don't have a positive control so I will need to investigate a bit more! Still not sure what causes this issue, perhaps a mismatch between all tiles in window vs. tiles with coefficients?

[snehamitra]

Hi @gprahalad ,

Is the ArchR peaks h5ad file a peak matrix where the peak coordinates are a subset of regions around the gene body? Since SCARlink works with the tile matrix, the plotting function is probably expecting values for all tiles around the gene body that led to a shape mismatch error.