Fail during numbat run : object 'seg' not found

[kokonech]

Error message is the following:

Testing for multi-allelic CNVs ..
0 multi-allelic CNVs found:
Evaluating CNV per cell ..
Mem used: 5.9Gb
Error in `mutate()`:
! Problem while computing `seg = factor(seg, mixedsort(unique(seg)))`.
Caused by error in `factor()`:
! object 'seg' not found

Settings and environment same as in issue #79
Attached produced figure.

[teng-gao]

Hi @kokonech ,

Thanks for finding this bug. This error was actually due to 1 cell having 0 UMI counts after filtering for mutually expressed genes between the reference and observation. The main github branch now fixes this; you can install via devtools.

Best,
Teng

[LouisFaure]

Hi @teng-gao, I am having a similar issue with a specific dataset, and this issue depends on which reference I use.
What I have is the following:

Error in `mutate()`:
ℹ In argument: `seg = factor(seg, mixedsort(unique(seg)))`.
Caused by error:
! object 'seg' not found

Could this be again an issue with a cell having 0 UMI? What would be the best way to check this?

Here is my last log:

INFO [2024-02-25 08:58:24]
Numbat version: 1.3.2.1
Scistreer version: 1.2.0
Running under parameters:
t = 1e-05
alpha = 1e-04
gamma = 20
min_cells = 50
init_k = 3
max_cost = 7765.8
n_cut = 0
max_iter = 2
max_nni = 100
min_depth = 0
use_loh = auto
segs_loh = None
call_clonal_loh = FALSE
segs_consensus_fix = None
multi_allelic = TRUE
min_LLR = 5
min_overlap = 0.45
max_entropy = 0.5
skip_nj = TRUE
diploid_chroms = None
ncores = 20
ncores_nni = 20
common_diploid = TRUE
tau = 0.3
check_convergence = FALSE
plot = TRUE
genome = hg38
Input metrics:
25886 cells
INFO [2024-02-25 08:58:25] Mem used: 11.3Gb
INFO [2024-02-25 08:58:46] Approximating initial clusters using smoothed expression ..
INFO [2024-02-25 08:58:48] Mem used: 11.3Gb
INFO [2024-02-25 09:23:46] running hclust...
INFO [2024-02-25 09:38:19] Iteration 1
INFO [2024-02-25 09:38:22] Mem used: 31.1Gb
INFO [2024-02-25 09:38:34] Expression noise level (MSE): medium (1.5).
INFO [2024-02-25 09:38:35] Running HMMs on 5 cell groups..
INFO [2024-02-25 09:38:55] diploid regions: 16b,16d
INFO [2024-02-25 09:42:02] Running HMMs on 3 cell groups..
INFO [2024-02-25 09:42:22] diploid regions: 16b,16d
INFO [2024-02-25 09:42:43] less than 5% of genome is in neutral region - including LOH in baseline
INFO [2024-02-25 09:43:50] Testing for multi-allelic CNVs ..
INFO [2024-02-25 09:43:50] 0 multi-allelic CNVs found:
INFO [2024-02-25 09:43:50] Evaluating CNV per cell ..
INFO [2024-02-25 09:43:52] Mem used: 15.5Gb
INFO [2024-02-25 09:44:07] less than 5% genes are in neutral region - including LOH in baseline
WARN [2024-02-25 10:51:27] 25886 cell(s) failed
WARN [2024-02-25 10:51:27] list("Error in summarise(., n = n(), phi_mle = calc_phi_mle_lnpois(Y_obs, lambda_ref,  : \n  \033[1m\033[22m\033[36mℹ\033[39m In argument: `phi_mle = calc_phi_mle_lnpois(...)`.\n\033[36mℹ\033[39m In group 2: `CHROM = 4`, `seg = \"4c\"`, `cnv_state = \"del\"`.\n\033[1mCaused by error in `solve.default()`:\033[22m\n\033[33m!\033[39m Lapack routine dgesv: system is exactly singular: U[1,1] = 0\n")
WARN [2024-02-25 10:51:27] KG66_N_AAACCCAAGCCTCACG-1