problem: dec_cci and dec_cci_all

[jinyuanchun]

Hi, When I used dec_cci and dec_cci_all to analysis my data about cell-cell communications
and it pops errors
dec_cci error ：

Error in intI(j, n = d[2L], dn[[2L]], give.dn = FALSE): invalid character indexing
Traceback:

1. dec_cci(object = obj, celltype_sender = "Vascular", celltype_receiver = "Differentiated_like", 
 .     use_n_cores = 4)
2. .lr_distance_doParallel(st_data, cell_pair, lrdb, celltype_sender, 
 .     celltype_receiver, per_num, pvalue)
3. st_data[lrdb$ligand, cell_pair$cell_sender]
4. st_data[lrdb$ligand, cell_pair$cell_sender]
5. callGeneric(x, i = i, j = j, drop = TRUE)
6. eval(call, parent.frame())
7. eval(call, parent.frame())
8. x[i = i, j = j, drop = TRUE]
9. x[i = i, j = j, drop = TRUE]
10. subCsp_ij(x, i, j, drop = drop)
11. intI(j, n = d[2L], dn[[2L]], give.dn = FALSE)
12. stop("invalid character indexing")

dec_cci_all error
Note: there are 6 cell types and 30 pair-wise cell pairs
Begin to find LR pairs
Error in {: task 1 failed - "invalid character indexing"
Traceback:

1. dec_cci_all(object = obj, use_n_cores = 4)
2. foreach::foreach(i = 1:nrow(celltype_pair), .packages = c("Matrix",
   . "reshape2"), .export = c(".get_cellpair", ".lr_distance",
   . ".get_tf_res", ".get_score")) %dopar% {
   . celltype_sender <- celltype_pair$celltype_sender[i]
   . celltype_receiver <- celltype_pair$celltype_receiver[i]
   . cell_pair <- .get_cellpair(celltype_dist, st_meta, celltype_sender,
   . celltype_receiver, n_neighbor)
   . cell_sender <- st_meta[st_meta$celltype == celltype_sender,
   . ]
   . cell_receiver <- st_meta[st_meta$celltype == celltype_receiver,
   . ]
   . cell_pair_all <- nrow(cell_sender) * nrow(cell_receiver)/2
   . if (nrow(cell_pair) > min_pairs) {
   . if (nrow(cell_pair) > cell_pair_all * min_pairs_ratio) {
   . lrdb <- object@lr_path$lrpairs
   . lrdb <- .lr_distance(st_data, cell_pair, lrdb, celltype_sender,
   . celltype_receiver, per_num, pvalue)
   . max_hop <- object@para$max_hop
   . receptor_tf <- NULL
   . if (nrow(lrdb) > 0) {
   . receptor_tf <- .get_tf_res(celltype_sender, celltype_receiver,
   . lrdb, ggi_tf, cell_pair, st_data, max_hop,
   . co_exp_ratio)
   . if (!is.null(receptor_tf)) {
   . lrdb <- .get_score(lrdb, receptor_tf)
   . }
   . else {
   . lrdb <- NULL
   . }
   . }
   . if (is.data.frame(lrdb)) {
   . if (nrow(lrdb) > 0) {
   . list(lrdb = lrdb, receptor_tf = receptor_tf,
   . cell_pair = cell_pair)
   . }
   . }
   . }
   . }
   . }
3. e$fun(obj, substitute(ex), parent.frame(), e$data)

[multitalk]

You can set if_doParallel = FALSE. I will check it later.

[jinyuanchun]

thank you for you reply. i try, but also have error.

[multitalk]

@jinyuanchun I have checked the code and I guess the reason might be that the nrow of obj@lr_path$lrpairs is equal to 1, which will lead to the error when calculating the .lr_distance(). I have fixed the bug considering the above condition. You can try it now. If the error still occurs, please feel free to contact me. Thanks again for your reply.

[multitalk]

@jinyuanchun Another reason is that the names of cells or spots contain "-". I have added a step to rename the cells or spots in st_data and sc_data over createSpaTalk() and dec_celltype by replacing "-" with "_" to avoid the error in dec_cci() and dec_cci_all(). Thanks again

[hahia]

@jinyuanchun Another reason is that the names of cells or spots contain "-". I have added a step to rename the cells or spots in st_data and sc_data over createSpaTalk() and dec_celltype by replacing "-" with "_" to avoid the error in dec_cci() and dec_cci_all(). Thanks again

Hi, I also meet the problem

names of st_data and names of celltype:

obj@lr_path$lrpairs:

[multitalk]

@hahia I guess it might be the names of obj@meta$rawmeta$cell. You could paste a C to the cellname, e.g., C111300, C30834, and re-run the codes starting from createSpaTalk()

[hahia]

@hahia I guess it might be the names of obj@meta$rawmeta$cell. You could paste a C to the cellname, e.g., C111300, C30834, and re-run the codes starting from createSpaTalk()

Ok! Problem solved.

[jinyuanchun]

thank you | | ***@***.*** | | ***@***.*** |

…

---- Replied Message ---- | From | ***@***.***> | | Date | 02/17/2023 09:31 | | To | ***@***.***> | | Cc | ***@***.***>***@***.***> | | Subject | Re: [ZJUFanLab/SpaTalk] problem: dec_cci and dec_cci_all (Issue #10) | @hahia I guess it might be the names of ***@***.***$rawmeta$cell. You could paste a C to the cellname, e.g., C111300, C30834, and re-run the codes starting from createSpaTalk() Ok! Problem solved. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: ***@***.***>

[liangbolyu]

Hi here is a question when I use the function dec_cci_all(). I'm doing a simulation to test the method, so I just have 2 cell types and two pathways. But I face the problem here:

Could you please help me with it? Thank you.