jayendrashinde91
commented
5 years ago Hi! Thank you for your query. Here is an example implementation of deconvolution_nmf(): denovo_signatures <- deconvolution_nmf(input_data = propSVsByCat, type = "SV", range_of_sigs = 2:12, nrun =20,method = "brunet",resdir = resdir.) The argument 'range_of_sigs' requires a range of possible number of mutational signatures (2:12) to estimate the most stable number of signatures.
If you already have established the exact estimate of the number of signatures to extract from your input data, the implementation would be as follows: denovo_signatures <- deconvolution_nmf(input_data = propSVsByCat, type = "SV", num_of_sigs = 12, nrun =20,method = "brunet",resdir = resdir.) The argument 'num_of_sigs' takes in the exact number of estimated signatures you provide. In your case nrun =50 & num_of_sigs = 12.
Best regards, Jay
AndreVidas
Hi Palimpsest team.
I'm working on the large 560 breast cancer data from Nik-Zainal et al. (https://doi.org/10.1038/nature17676) in the pursuit of reproducing their rearrangement signatures. I'm able to get the signatures by setting nrun = 30 & range_of_sigs = 12. However in this process I noticed a lot of errors from other runs with the same data but with other settings of nrun and range_of_sigs:
When changing NRUN
nrun =50 & range_of_sigs = 12
[1] "Estimating the optimal number of mutational signatures..."
Timing stopped at: 193.7 38.88 318.3 Timing stopped at: 158.8 29.58 334.8 Timing stopped at: 205.5 40.71 419.8 Timing stopped at: 285.6 44.92 375 Timing stopped at: 252.6 35.51 507.2 Timing stopped at: 402.6 49.61 454 Timing stopped at: 396.7 52.06 537.6 Timing stopped at: 453.8 49.97 563.3 Timing stopped at: 434.3 50.08 532.1 Timing stopped at: 264.8 44.34 374.2 Timing stopped at: 394.1 39.35 707.8 Error in (function (...) : All the runs produced an error: -#1 [r=2] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#2 [r=3] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#3 [r=4] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#4 [r=5] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#5 [r=6] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#6 [r=7] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#7 [r=8] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#8 [r=9] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#9 [r=10] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#10 [r=11] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#11 [r=12] -> (list) object cannot be coerced to type 'double' [in call t Calls: deconvolution_nmf -> nmfEstimateRank -> do.call ->
Execution halted
nrun = 70 & range_of_sigs = 12
[1] "Estimating the optimal number of mutational signatures..."
Timing stopped at: 142.4 39.54 239 Timing stopped at: 220.7 36.57 334.9 Timing stopped at: 245.6 45.64 367.7 Timing stopped at: 215.7 37.63 372.5 Timing stopped at: 387 42.69 504.4 Timing stopped at: 283.5 43 395.7 Timing stopped at: 426.1 50.33 558.4 Timing stopped at: 395.5 48.59 528.5 Timing stopped at: 575.1 50.45 626.4 Timing stopped at: 228.7 39.92 312.6 Timing stopped at: 375 44.68 539.6 Error in (function (...) : All the runs produced an error: -#1 [r=2] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#2 [r=3] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#3 [r=4] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#4 [r=5] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#5 [r=6] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#6 [r=7] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#7 [r=8] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#8 [r=9] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#9 [r=10] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#10 [r=11] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#11 [r=12] -> (list) object cannot be coerced to type 'double' [in call t Calls: deconvolution_nmf -> nmfEstimateRank -> do.call ->
Execution halted
nrun = 80 & range_of_sigs = 12
[1] "Calculating exposures of signatures in the input tumors"
Timing stopped at: 117.7 22.11 426.2 Timing stopped at: 188 35.73 410.3 Timing stopped at: 289.2 35.57 496.8 Timing stopped at: 241.9 41.26 342.2 Timing stopped at: 482.6 48.48 632.1 Timing stopped at: 234.3 49.22 290.3 Timing stopped at: 273.6 30.33 814 Timing stopped at: 179 41.72 303.1 Timing stopped at: 454.7 40.21 905.5 Timing stopped at: 759.5 52.74 1031 geom_path: Each group consists of only one observation. Do you need to adjust the group aesthetic? geom_path: Each group consists of only one observation. Do you need to adjust the group aesthetic? Error: NMF::nmf - invalid argument 'rank': must be a single numeric value In addition: Warning messages: 1: In (function (...) : NAs were produced due to errors in some of the runs: -#1[r=2] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#2[r=3] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#3[r=4] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#4[r=5] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#5[r=6] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#6[r=7] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#7[r=8] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#8[r=9] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#9[r=10] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#10[r=11] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] 2: Removed 40 rows containing missing values (geom_path). 3: Removed 100 rows containing missing values (geom_point). 4: In max(abs(diff(z))) : no non-missing arguments to max; returning -Inf Execution halted
When changing RANGE_OF_SIGS
nrun =60 & range_of_sigs = 8
[1] "Estimating the optimal number of mutational signatures..."
Timing stopped at: 219.6 40.97 309.6 Timing stopped at: 290.9 42.98 352 Timing stopped at: 243 35.3 454.1 Timing stopped at: 131.2 20.06 530.5 Timing stopped at: 224.1 28.04 671.6 Timing stopped at: 301.6 38.47 530.2 Timing stopped at: 545.8 52.13 795.9 Error in (function (...) : All the runs produced an error: -#1 [r=2] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#2 [r=3] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#3 [r=4] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#4 [r=5] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#5 [r=6] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#6 [r=7] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] -#7 [r=8] -> (list) object cannot be coerced to type 'double' [in call to 'which.min'] Calls: deconvolution_nmf -> nmfEstimateRank -> do.call ->
Execution halted
Do you have any idea of why this is happening?
Best regards, André
sessionInfo()
R version 3.5.0 (2018-04-23) Platform: x86_64-pc-linux-gnu (64-bit) Running under: CentOS Linux 7 (Core)
Matrix products: default BLAS/LAPACK: /cm/shared/apps/intel/parallel_studio_xe/2018_update2/compilers_and_libraries_2018.2.199/linux/mkl/lib/intel64_lin/libmkl_gf_lp64.so
locale: [1] C
attached base packages: [1] stats4 parallel stats graphics grDevices utils datasets [8] methods base
other attached packages: [1] getopt_1.20.2 Palimpsest_1.0.0 GenomicRanges_1.32.7 [4] GenomeInfoDb_1.16.0 IRanges_2.14.12 S4Vectors_0.18.3
[7] BiocInstaller_1.30.0 NMF_0.21.0 bigmemory_4.5.33
[10] Biobase_2.40.0 BiocGenerics_0.26.0 cluster_2.0.7-1
[13] rngtools_1.3.1 pkgmaker_0.27 registry_0.5
[16] RColorBrewer_1.1-2 RCircos_1.2.0 bedr_1.0.4
[19] RLinuxModules_0.2
loaded via a namespace (and not attached): [1] lsa_0.73.1 bitops_1.0-6
[3] matrixStats_0.54.0 bit64_0.9-7
[5] doParallel_1.0.14 progress_1.2.0
[7] httr_1.3.1 Rgraphviz_2.24.0
[9] SnowballC_0.5.1 tools_3.5.0
[11] R6_2.3.0 KernSmooth_2.23-15
[13] DBI_1.0.0 lazyeval_0.2.1
[15] colorspace_1.3-2 withr_2.1.2
[17] tidyselect_0.2.5 prettyunits_1.0.2
[19] bit_1.1-14 compiler_3.5.0
[21] VennDiagram_1.6.20 graph_1.58.2
[23] formatR_1.5 DelayedArray_0.6.6
[25] rtracklayer_1.40.6 caTools_1.17.1.1
[27] scales_1.0.0 stringr_1.3.1
[29] digest_0.6.18 Rsamtools_1.32.3
[31] R.utils_2.7.0 XVector_0.20.0
[33] pkgconfig_2.0.2 bibtex_0.4.2
[35] plotrix_3.7-4 BSgenome_1.48.0
[37] rlang_0.3.0.1 RSQLite_2.1.1
[39] bindr_0.1.1 gtools_3.8.1
[41] BiocParallel_1.14.2 dplyr_0.7.7
[43] R.oo_1.22.0 VariantAnnotation_1.26.1
[45] RCurl_1.95-4.11 magrittr_1.5
[47] GenomeInfoDbData_1.1.0 futile.logger_1.4.3
[49] Matrix_1.2-15 Rcpp_1.0.0
[51] munsell_0.5.0 R.methodsS3_1.7.1
[53] stringi_1.2.4 yaml_2.2.0
[55] SummarizedExperiment_1.10.1 zlibbioc_1.26.0
[57] gplots_3.0.1 plyr_1.8.4
[59] grid_3.5.0 blob_1.1.1
[61] gdata_2.18.0 bigmemory.sri_0.1.3
[63] crayon_1.3.4 lattice_0.20-38
[65] Biostrings_2.48.0 GenomicFeatures_1.32.3
[67] hms_0.4.2 pillar_1.3.0
[69] reshape2_1.4.3 codetools_0.2-15
[71] biomaRt_2.36.1 futile.options_1.0.1
[73] XML_3.98-1.16 glue_1.3.0
[75] lambda.r_1.2.3 data.table_1.11.8
[77] foreach_1.4.4 testthat_2.0.1
[79] gtable_0.2.0 purrr_0.2.5
[81] assertthat_0.2.0 ggplot2_3.1.0
[83] gridBase_0.4-7 xtable_1.8-3
[85] tibble_1.4.2 iterators_1.0.10
[87] GenomicAlignments_1.16.0 AnnotationDbi_1.42.1
[89] memoise_1.1.0 bindrcpp_0.2.2