The output file "segmentation.csv" cell column value is NaN

[mzj0418]

Hello!

Thank you for the tool. I am interested in trying this tool on my own data(Stereo-seq data['geneID', 'x', 'y', 'MIDCounts']).

baysor run ./stereo-seq.csv -p -c ./configs/example_config.toml -p -o ./stereo-seq_output_no_dapi/ -m 1 -g 'geneID' -s 8
[17:11:06] Info: Run R1b9f25c4f
[17:11:06] Info: (2023-09-15) Run Baysor v0.6.2
[17:11:06] Info: Loading data...
[17:11:08] Info: Loaded 138957 transcripts
[17:11:10] Info: Estimating noise level
[17:11:16] Info: Done
[17:11:29] Info: Clustering molecules...
Progress: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| Time: 0:00:02
Iteration: 66
Max. difference: 9.26e-5
Fraction of probs changed: 0.00238
[17:12:51] Info: Algorithm stopped after 66 iterations. Error: 9.26e-5. Converged: true.
[17:12:51] Info: Done
[17:12:51] Info: Initializing algorithm. Scale: 8.0, scale std: 2.0, initial #components: 92638, #molecules: 138957.
[17:12:51] Warning: n=92638, which is > length(high_conf_ids) (2553)
└ Baysor.Processing /home/viktor_petukhov/.julia/dev/Baysor/src/processing/data_processing/initialization.jl:180
[17:12:52] Info: Using the following additional information about molecules: [:confidence, :cluster]
[17:12:52] Info: Using 2D coordinates
Progress: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| Time: 0:03:45
Iteration: 500
Noise level, %: 98.17
Num. components: 319
[17:16:39] Info: Processing complete.
[17:16:39] Info: Estimating local colors
[17:18:42] Warning: n=10000, which is > length(high_conf_ids) (2553)
└ Baysor.Processing /home/viktor_petukhov/.julia/dev/Baysor/src/processing/data_processing/initialization.jl:180
[17:24:19] Info: Estimating boundary polygons
[17:24:20] Info: Saving results to ./stereo-seq_output_no_dapi/segmentation.csv
Progress: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| Time: 0:00:07
[17:24:35] Info: Plotting results
[17:24:37] Info: Plot diagnostics
[17:24:39] Info: Plot transcript assignment
[17:24:46] Info: All done!
(process:3486428): GLib-CRITICAL **: 17:24:46.069: g_datalist_id_set_data_full: assertion 'key_id > 0' failed
(process:3486428): GLib-CRITICAL **: 17:24:46.069: g_datalist_id_set_data_full: assertion 'key_id > 0' failed
(process:3486428): GLib-CRITICAL **: 17:24:46.070: g_datalist_id_set_data_full: assertion 'key_id > 0' failed

I don't know if this is correct, I looked at the output file "segmentation.csv" and found that the cell column value is NaN, can you help me find this error?

Best,
Jun

[Anne-LiseGerard]

Hi Jun, I run into a similar issue when running baysor v0.7.1 on an HPC. I get the exact same GLib-CRITICAL message:

[...]
[03:15:33] Info: Saving results to /g/data/ug9/Immune/ag3362/segmentation.csv
[03:21:03] Info: Plotting results
[03:21:05] Info: Plot diagnostics
[03:21:26] Info: Plot transcript assignment
[03:29:12] Info: All done!
(process:3820376): GLib-CRITICAL **: 03:21:30.294: g_datalist_id_set_data_full: assertion 'key_id > 0' failed
(process:3820376): GLib-CRITICAL **: 03:29:12.240: g_datalist_id_set_data_full: assertion 'key_id > 0' failed
(process:3820376): GLib-CRITICAL **: 03:29:12.241: g_datalist_id_set_data_full: assertion 'key_id > 0' failed

However, my segmentation.csv file does contain values in the 'cell' column whenever the 'is_noise' value is FALSE.

@mzj0418 did you end up solving this?
@VPetukhov can we trust baysor outputs when this error is returned (so far as I can tell, the output files look good)? If not, is there something that can be done?

Thank you in advance for your help!
Anne-Lise

PS: looking at the timestamps, something seems to be going wrong during the final plotting steps, if that helps.