Closed Pointillomic closed 1 year ago
Hi @Pointillomic,
We're very glad to hear that Giotto is helpful!
I tried taking a look at this issue and was able to confirm for FOVs 31 and 32 in Lung5rep1, but it's a little strange. The information from fovs 31 and 32 seem to be absent from metadata_file.csv
and tx_file.csv
even though there are entries for them in the fov_positions_file.csv
and images for them in all the image folders.
This lack of data results in unexpectedly empty data.tables that result in the error message that you found.
I tried looking for hints in the feature detections in case they were accidentally mis-categorized into a different FOV in the feature detections file, but I didn't manage to find anything out of the ordinary.
tx_path = '/path/to/Lung5_Rep1_tx_file.csv'
tx_coord_all = data.table::fread(input = tx_path)
fov_counts = tx_coord_all[, table(fov)]
barplot(fov_counts)
Detections per FOV: nothing really stands out as also having additional detections that should have been in 31 and 32.
local_max = tx_coord_all[, lapply(.SD, max), by = fov, .SDcols = c('x_local_px', 'y_local_px')]
plot(x = local_max$x_local_px, y = local_max$y_local_px, xlim = c(5100, 5900), ylim = c(3200, 4000))
Local max x and y values (max values relative to FOV for the detection points) are all very similar for each FOV
global_mean = tx_coord_all[, lapply(.SD, max), by = fov, .SDcols = c('x_global_px', 'y_global_px')]
plot(x = global_mean$x_global_px, y = global_mean$y_global_px)
Mean detection location for each FOV is regularly spaced for all 30 FOVs in the data
meta_path = '/path/to/Lung5_Rep1_metadata_file.csv'
meta = data.table::fread(meta_path)
meta[, unique(fov)]
[1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
[17] 17 18 19 20 21 22 23 24 25 26 27 28 29 30
The data for 31 and 32 also seem to be missing from the aggregate information
Best, Jiaji
Hi Jiaji,
Thank you so much for looking into this so quickly!
This is super helpful since this indicates a problem with the data rather than the package, so I can go ahead with the analysis excluding those problematic FOV
Thanks again for the great package!
Hello,
First of all, thank you for the amazing tool! It has made spatial transcriptomic analysis with the CosMX so much easier!
I had an issue with the last two versions of/commits to Giotto suite with loading specific FOV from the public lung CosMX dataset. Specifically, I would get the error "Error: [$] feat_ID is not a variable name in x", when loading FOV 31 and 32 for lung5_Rep1, whereas it works for the other FOV for that sample. Because it is so FOV based, I tried redownloading and checking the md5sum, but it seems to be fine.
The specific function that has the issue is 'createGiottoObjectSubcellular' and I included the output just before the error below.
Any help with this would be greatly appreciated, and thanks again for the great tool!
$mask_method [1] "guess"
$flip_vertical [1] TRUE
$flip_horizontal [1] FALSE
$shift_horizontal_step [1] FALSE
Selecting col "geom" as poly_ID column Selecting cols "x" and "y" as x and y respectively [1] 0 [1] 3648