Issue with Startrac.run(): dcast Error in pIndex.migr Calculation

[s4rivera]

Hello! I am really looking forward to using your very exciting tool!

Currently I am having the following issue with my data which I confirmed when trying to reproduce your vignette:
https://htmlpreview.github.io/?https://github.com/Japrin/STARTRAC/blob/master/vignettes/startrac.html

I am encountering an issue when trying to run Startrac.run() using the provided example dataset (example.cloneDat.Zhang2018.txt). The pipeline fails during the pairwise index calculation with an error related to dcast from data.table.

System Information:

R version: 4.4.2 (2024-10-31)
Startrac version: 0.1.0
data.table version: 1.17.0
Reproducible Example:
I followed the Startrac vignette and ran the following minimal script:

library(Startrac)
library(data.table)
library(dplyr)

Define file path

file_path <- "example.cloneDat.Zhang2018.txt"

Read the file

in.dat <- fread(file_path, stringsAsFactors = FALSE, header = TRUE)

Run Startrac pipeline

tic("Startrac.run")
out <- Startrac.run(in.dat, proj="CRC", verbose=F)
Error Message:

[2025-03-17 16:34:42.528854] initialize Startrac ...
[2025-03-17 16:34:42.557158] calculate startrac index ...
[2025-03-17 16:34:42.577406] calculate pairwise index ...

Error in dcast(cls.migr.index.df, majorCluster ~ crossLoc, value.var = "pIndex.migr") : The dcast generic in data.table has been passed a data.frame, but data.table::dcast currently only has a method for data.tables. Please confirm your input is a data.table, with setDT(cls.migr.index.df) or as.data.table(cls.migr.index.df). If you intend to use a method from reshape2, try installing that package first, but do note that reshape2 is superseded and is no longer actively developed.

What I’ve Tried:

Confirmed that the input dataset is correctly formatted
Verified column names match expected names (Cell_Name, clone.id, patient, majorCluster, loc).
Checked the class of cls.migr.index.df
It appears to be a data.frame, but dcast() expects a data.table.
Tried setDT(cls.migr.index.df) before running pIndex(), but it did not resolve the issue.
Tried running pIndex(obj) separately, same error persists.
This issue persists even when using the example dataset. Is there a known fix for this, or should Startrac.run() be handling this conversion internally? Any help would be appreciated!

Thanks so much!

Sam

[Minazukiz]

on the error message, I guessed that it was related to the version of data.table, so I checked the environment configuration provided by the author:

remotes::install_version("data.table", version = "1.12.2")
> out2 <- Startrac.run(subT.meta, proj="CRC",verbose=F)
[2025-03-25 16:53:49.357245] initialize Startrac ...
[2025-03-25 16:53:49.423915] calculate startrac index ...
[2025-03-25 16:53:49.446715] calculate pairwise index ...
[2025-03-25 16:53:49.996739] calculate indices of each patient ...
[2025-03-25 16:53:50.223995] collect result
[2025-03-25 16:53:50.226733] return

seems work and returns results, but I still don't understand the root cause of this error

[xutongran]

Hello! I am really looking forward to using your very exciting tool!

Currently I am having the following issue with my data which I confirmed when trying to reproduce your vignette: https://htmlpreview.github.io/?https://github.com/Japrin/STARTRAC/blob/master/vignettes/startrac.html

I am encountering an issue when trying to run Startrac.run() using the provided example dataset (example.cloneDat.Zhang2018.txt). The pipeline fails during the pairwise index calculation with an error related to dcast from data.table.

System Information:

R version: 4.4.2 (2024-10-31) Startrac version: 0.1.0 data.table version: 1.17.0 Reproducible Example: I followed the Startrac vignette and ran the following minimal script:

library(Startrac) library(data.table) library(dplyr)

Define file path

file_path <- "example.cloneDat.Zhang2018.txt"

Read the file

in.dat <- fread(file_path, stringsAsFactors = FALSE, header = TRUE)

Run Startrac pipeline

tic("Startrac.run") out <- Startrac.run(in.dat, proj="CRC", verbose=F) Error Message:

[2025-03-17 16:34:42.528854] initialize Startrac ... [2025-03-17 16:34:42.557158] calculate startrac index ... [2025-03-17 16:34:42.577406] calculate pairwise index ...

Error in dcast(cls.migr.index.df, majorCluster ~ crossLoc, value.var = "pIndex.migr") : The dcast generic in data.table has been passed a data.frame, but data.table::dcast currently only has a method for data.tables. Please confirm your input is a data.table, with setDT(cls.migr.index.df) or as.data.table(cls.migr.index.df). If you intend to use a method from reshape2, try installing that package first, but do note that reshape2 is superseded and is no longer actively developed.
What I’ve Tried:

Confirmed that the input dataset is correctly formatted Verified column names match expected names (Cell_Name, clone.id, patient, majorCluster, loc). Checked the class of cls.migr.index.df It appears to be a data.frame, but dcast() expects a data.table. Tried setDT(cls.migr.index.df) before running pIndex(), but it did not resolve the issue. Tried running pIndex(obj) separately, same error persists. This issue persists even when using the example dataset. Is there a known fix for this, or should Startrac.run() be handling this conversion internally? Any help would be appreciated!

Thanks so much!

Sam

I'm encountering the same error with Startrac. Have you resolved this issue now?

[lisch7]

devtools::install_version("data.table", version = "1.16.2") works for me