Errors in "delta_auc_in_pseudospace" function.

[SiyiWanggou]

Hi Team from cole-trapnell-lab,
I am using your wonderful piplines for trajectory comparison. It is really awesome and I love it so much!
Recently, I meet some errors when I use "delta_auc_in_pseudospace" function in "cds_exprs.auc" calculation. The system reports the Error like this "This tidyselect interface doesn't support predicates yet". I checked the funtion code and just wonder whether the "id" should be "gene_id" here. The original code is listed below:

cds_exprs.auc = cds_exprs %>% dplyr::select(id, Cell.Type,
Pseudotime, expectation) %>% dplyr::mutate(pseudotime_quantile = cut(Pseudotime,
breaks = seq(0, pseudotime_max, pseudotime_max/5))) %>%
unique() %>% dplyr::group_by(id, Cell.Type, pseudotime_quantile) %>%
dplyr::summarize(AUC = MESS::auc(Pseudotime, expectation,
type = "spline")) %>% dplyr::group_by(id, pseudotime_quantile) %>%
dplyr::summarize(auc_difference = (AUC[1] - AUC[2])/(AUC[1] +
AUC[2])) %>% dplyr::arrange(desc(auc_difference))

Thanks,
Siyi

[joselmc]

Hi Siyi, At that point during the function cds_exprs should contain both a gene_id and an id column due to an earlier merge. I assume the error is prior to the dplyr call, likely before the merge step, and probably due to changes or conflicts in a newer version of a package. Could you print your sessionInfo() ?