error when using interaction term in formla, in ancombc2

[boaty]

Dear author,

I found a error when running anbombc2 with interaction term in formula.

ancombc2 version : 2.4.0

my R codes
out = ancombc2(data =phy_tmp, tax_level = "Species", fix_formula = "TimePoint * Group + AGE + SEX", rand_formula = "(1 + TimePoint |siteID) + ( 1 | siteID:subjectID)", p_adj_method = "BH", pseudo_sens = TRUE, prv_cut = 0.10, lib_cut = 1000, s0_perc = 0.05, group = "Group", struc_zero = TRUE, neg_lb = TRUE, alpha = 0.05, n_cl = 5, verbose = TRUE, global = T, pairwise = T, dunnet = T, trend = F, iter_control = list(tol = 1e-2, max_iter = 20, verbose = TRUE), em_control = list(tol = 1e-5, max_iter = 100), lme_control = lme4::lmerControl(), mdfdr_control = list(fwer_ctrl_method = "BH", B = 100) )

error message:
Error in .data_qc(meta_data = core2$meta_data, formula = fix_formula, : The following variables specified are not in the meta data: TimePoint * Group

So i did some dig in, found the function .data_qc() should take fix_formula as a reference to check the metadata varaible, but somehow it can not recognize "*" .

I replaced "TimePoint * Group + AGE + SEX" by "TimePoint + Group + AGE + SEX" for .data_qc() function. And use original "TimePoint * Group + AGE + SEX" for later modeling. i run as expected.

So, is it a bug or you did this for some reasons to disable interaction terms?

thanks

Zhou

[jamesriddellv]

Running into the same issue with 2.6.0.

Error in .data_qc(meta_data = core2$meta_data, formula = fix_formula, :
The following variables specified are not in the meta data: timepoint * treatment

I also ran on the test code from the ANCOM-BC2 tutorial and encounter the same error:

`> library(ANCOMBC)

library(dplyr)
library(phyloseq)
library(microbiome)
library(readxl)
data(dietswap, package = "microbiome")
data(dietswap, package = "microbiome")
output = ancombc2(data = dietswap, tax_level = "Family",
fix_formula = "nationality + timepoint * group",
rand_formula = "(timepoint | subject)",
p_adj_method = "holm", pseudo_sens = TRUE,
prv_cut = 0.10, lib_cut = 1000, s0_perc = 0.05,
group = "group", struc_zero = TRUE, neg_lb = TRUE,
alpha = 0.05, n_cl = 2, verbose = TRUE,
global = TRUE, pairwise = TRUE, dunnet = TRUE, trend = TRUE,
iter_control = list(tol = 1e-2, max_iter = 20,
verbose = TRUE),
em_control = list(tol = 1e-5, max_iter = 100),
lme_control = lme4::lmerControl(),
mdfdr_control = list(fwer_ctrl_method = "holm", B = 100),
trend_control = list(contrast = list(matrix(c(1, 0, -1, 1),
nrow = 2,
byrow = TRUE)),
node = list(2),
solver = "ECOS",
B = 10))
Error in .data_qc(meta_data = core2$meta_data, formula = fix_formula, :
The following variables specified are not in the meta data: timepoint * group`

SessionInfo()
R version 4.4.0 (2024-04-24)
Platform: aarch64-apple-darwin20
Running under: macOS 15.1.1

Matrix products: default
BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.4-arm64/Resources/lib/libRlapack.dylib; LAPACK version 3.12.0

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

time zone: America/Detroit
tzcode source: internal

attached base packages:
[1] stats4 stats graphics grDevices utils datasets methods
[8] base

other attached packages:
[1] microbiome_1.26.0 ANCOMBC_2.6.0
[3] vegan3d_1.3-0 readxl_1.4.3
[5] corrplot_0.95 apcluster_1.4.13
[7] lubridate_1.9.3 forcats_1.0.0
[9] stringr_1.5.1 purrr_1.0.2
[11] readr_2.1.5 tibble_3.2.1
[13] tidyverse_2.0.0 BiocManager_1.30.25
[15] tidyr_1.3.1 dplyr_1.1.4
[17] ggplot2_3.5.1 viridis_0.6.5
[19] viridisLite_0.4.2 vegan_2.6-8
[21] lattice_0.22-6 permute_0.9-7
[23] phyloseq_1.48.0 TreeSummarizedExperiment_2.12.0
[25] Biostrings_2.72.1 XVector_0.44.0
[27] SingleCellExperiment_1.26.0 SummarizedExperiment_1.34.0
[29] Biobase_2.64.0 GenomicRanges_1.56.2
[31] GenomeInfoDb_1.40.1 IRanges_2.38.1
[33] S4Vectors_0.42.1 BiocGenerics_0.50.0
[35] MatrixGenerics_1.16.0 matrixStats_1.4.1

loaded via a namespace (and not attached):
[1] splines_4.4.0 cellranger_1.1.0
[3] rpart_4.1.23 DirichletMultinomial_1.46.0
[5] lifecycle_1.0.4 Rdpack_2.6.2
[7] doParallel_1.0.17 MASS_7.3-61
[9] MultiAssayExperiment_1.30.3 backports_1.5.0
[11] magrittr_2.0.3 rmarkdown_2.29
[13] Hmisc_5.2-1 doRNG_1.8.6
[15] gld_2.6.6 DBI_1.2.3
[17] minqa_1.2.8 ade4_1.7-22
[19] multcomp_1.4-26 abind_1.4-8
[21] zlibbioc_1.50.0 Rtsne_0.17
[23] expm_1.0-0 TH.data_1.1-2
[25] nnet_7.3-19 rgl_1.3.14
[27] yulab.utils_0.1.8 sandwich_3.1-1
[29] GenomeInfoDbData_1.2.12 ggrepel_0.9.6
[31] irlba_2.3.5.1 tidytree_0.4.6
[33] DelayedMatrixStats_1.26.0 codetools_0.2-20
[35] DelayedArray_0.30.1 energy_1.7-12
[37] scuttle_1.14.0 tidyselect_1.2.1
[39] UCSC.utils_1.0.0 farver_2.1.2
[41] lme4_1.1-35.5 gmp_0.7-5
[43] ScaledMatrix_1.12.0 base64enc_0.1-3
[45] jsonlite_1.8.9 BiocNeighbors_1.22.0
[47] multtest_2.60.0 e1071_1.7-16
[49] Formula_1.2-5 decontam_1.24.0
[51] mia_1.12.0 survival_3.7-0
[53] scater_1.32.1 iterators_1.0.14
[55] systemfonts_1.1.0 foreach_1.5.2
[57] tools_4.4.0 treeio_1.28.0
[59] ragg_1.3.3 DescTools_0.99.58
[61] Rcpp_1.0.13-1 glue_1.8.0
[63] gridExtra_2.3 SparseArray_1.4.8
[65] xfun_0.49 mgcv_1.9-1
[67] numDeriv_2016.8-1.1 withr_3.0.2
[69] fastmap_1.2.0 boot_1.3-31
[71] rhdf5filters_1.16.0 bluster_1.14.0
[73] fansi_1.0.6 digest_0.6.37
[75] rsvd_1.0.5 timechange_0.3.0
[77] R6_2.5.1 textshaping_0.4.1
[79] colorspace_2.1-1 gtools_3.9.5
[81] utf8_1.2.4 generics_0.1.3
[83] data.table_1.16.4 DECIPHER_3.0.0
[85] class_7.3-22 CVXR_1.0-15
[87] httr_1.4.7 htmlwidgets_1.6.4
[89] S4Arrays_1.4.1 scatterplot3d_0.3-44
[91] pkgconfig_2.0.3 gtable_0.3.6
[93] Exact_3.3 Rmpfr_1.0-0
[95] htmltools_0.5.8.1 biomformat_1.32.0
[97] scales_1.3.0 lmom_3.2
[99] knitr_1.49 rstudioapi_0.17.1
[101] tzdb_0.4.0 reshape2_1.4.4
[103] checkmate_2.3.2 nlme_3.1-166
[105] nloptr_2.1.1 zoo_1.8-12
[107] proxy_0.4-27 rhdf5_2.48.0
[109] rootSolve_1.8.2.4 parallel_4.4.0
[111] vipor_0.4.7 foreign_0.8-87
[113] pillar_1.9.0 grid_4.4.0
[115] vctrs_0.6.5 BiocSingular_1.20.0
[117] beachmat_2.20.0 cluster_2.1.6
[119] htmlTable_2.4.3 beeswarm_0.4.0
[121] evaluate_1.0.1 mvtnorm_1.3-2
[123] cli_3.6.3 compiler_4.4.0
[125] rngtools_1.5.2 rlang_1.1.4
[127] crayon_1.5.3 labeling_0.4.3
[129] plyr_1.8.9 fs_1.6.5
[131] ggbeeswarm_0.7.2 stringi_1.8.4
[133] BiocParallel_1.38.0 lmerTest_3.1-3
[135] gsl_2.1-8 munsell_0.5.1
[137] lazyeval_0.2.2 Matrix_1.7-1
[139] hms_1.1.3 sparseMatrixStats_1.16.0
[141] bit64_4.5.2 Rhdf5lib_1.26.0
[143] haven_2.5.4 rbibutils_2.3
[145] igraph_2.1.1 bit_4.5.0.1
[147] ape_5.8

[ctilves]

I was receiving the same errors, and realized that the version downloaded using BiocManager::install("ANCOMBC") was not actually the updated version (despite still saying it is version 2.6.0). Installing directly from github using devtools::install_github("FrederickHuangLin/ANCOMBC") seems to install the correct version 2.6.0, and I was able to run models including the interaction term.

[jamesriddellv]

Thanks @ctilves this fixed the issue for me!

[ScottWrigley]

I came here looking for a solution to this issue as well. Thank you, @ctilves. Your suggestion solved the problem, and the model will now run with interaction terms.