Fixed bug for signatureBoxplot by removing quotations around variables. This issue was introduced with the newest ggplot2 version. Thanks to Arthur VanValkenburg for identifying the issue and solution.
Removed Zimmer_RES_3 which was an identical signature as the previously published Sweeney_OD_3.
Fixed error in example for compare_algs caused by the signature being used.
Minor Changes
Added Vargas_18 and Vargas_45 signatures (doi: 10.1371/journal.pcbi.1010770)
TBSignatureProfiler 1.12.0
Bug Fixes
Removed pROC argument from TableAUC() as it was causing a bug where the upper CI was always the same as the AUC point estimate, even with pROC=FALSE as the default setting.
Changed source package for rowSums() and rowMeans() to base from BiocGenerics due to an error on the Bioconductor testing servers.
Minor Changes
Added Chen_5 signature
TBSignatureProfiler 1.10.0
Bug Fixes
Fixed the tableAUC bootstrapped confidence interval to be the 2.5 and 97.5 percentiles instead of the 5 and 95 percentiles
Fixed the upper CI value for the pROC/DeLong AUC CI method in the bootstrapAUC function
Major Changes
Changed tableAUC confidence interval default to bootstrapped instead of DeLong (pROC argument)
Minor Changes
Updated the github and website introductions.
Added Natarajan_7, Kaul_3 signatures
Added Francisco_OD_2, Kwan_186 signatures
Shortened some example run times
TBSignatureProfiler 1.8.0
Bug Fixes
Fixed gene in RESPONSE5 (PNN to RP11-295G20.2) in TBsignatures and TBcommon objects. (Stanley M. Kimbung)
Fixed a bug in the singscore algorithm called from runTBsigProfiler() that would not allow for the scoring of user-provided signatures.
Fixed the TB_hiv data to remove unnecessary factor level of Disease metadata.
Fixed the row numbers of existing sigAnnotData and common_sigAnnotData objects, and added code to update them after new signatures are added.
Major Changes
If any signatures in the object used with runTBsigProfiler() have <2 genes present in the given sample, the signatures will not be scored. This may affect existing scripts.
Minor Changes
Reorganized code for OriginalModel.R for clarity.
Added 4 new signatures (Tabone_OD_11/TB12, Tabone_RES_25/EarlyRESP-TB25, Tabone_RES_27/TREAT-TB27, Long_RES_10)
Updated the website interface
Changed HGNChelper installation to be checked during profiling if update_genes = TRUE
Reorganized code in mkAssay() for clarity. The output_name argument is now appended to all output assays, whereas previously it was only appended to the log of the input assay.
TBSignatureProfiler 1.6.0
Bug Fixes
Fixed some package tests for the profile.R, OriginalModel.R scripts to increase coverage
Major Changes
Added the gene_update parameter to runTBsigProfiler() to allow users to check signature/sample gene names for excel mogrified or outdated gene symbols
Updated gene symbols for original models functions
Added Chendi_HIV_2 signature (8/24)
Removed Mendelsoh_RISK_11 as it was originally published in Darboe_RISK_11. Changed Darboe_RISK_11 to RISK11 in common objects. (Thomas Scriba)
Minor Changes
Rewrote runTBsigProfiler() to be shorter and easier to maintain
TBSignatureProfiler 1.5.0
68 signatures currently available
Bug Fixes
Fixed incorrect names of signatures in Tbcommon and common_sigAnnotData objects
Major Changes
Added Zimmer_RES_3, Gong_OD_4, Bloom_RES_268, and Bloom_RES_558
Added Sivakumaran_11 and Mendelsoh_RISK_11 signatures
Added Estevez_133, Estevez_259, LauxdaCosta_OD_3, and Maertzdorf_15 signatures to the package
Added Chen_HIV_4, Gliddon_HIV_3, Gliddon_2_OD_4, Kulkarni_HIV_2, and Heycken_FAIL_22 signatures
Added a COVIDsignatures object to the package that can be used to profile COVID-19 gene transcript signatures, thanks to collaborator Dylan Sheerin (WEHI)
Added functions to evaluate some signatures using their original models from Johnson lab member Xutao Wang
Minor Changes
Added mention of COVIDsignatures object to main vignette and website
Included the OG models tutorial on website (Xutao Wang)
Added addTBsignature() to more easily facilitate updating signatures in package
Added pROC option to obtain confidence intervals on AUC values as part of