NEWS

---

scTensor 2.4.1

• A vignette modified.

scTensor 2.4.0

• Regularizer parameter (L2_A/L1_A) was added in cellCellDecomp() ("ntd", "ntd2").
• Multilinear CX Decompotision was added in cellCellDecomp() ("cx").
• convertNCBIGeneID is removed.
• The vignettes were modified.
• Support of LRBase.XXX.eg.db-type packages is completely deprecated

scTensor 2.0.0

• Extended to use the version 2.0.0 of LRBase.XXX.eg.db-type packages
• Omitted typing Enter-key many times to perform example('cellCellReport')
• lr.evidence option was added in cellCellRanks() and cellCellDecomp() to select ligand-receptor databases to construct CCI-tensor (cf. Evidence code: https://github.com/rikenbit/lrbase-workflow)
• The L-R evidence was embeded in the HTML report.
• The bug related to the hyper-link was fixed in .hyperLinks.
• Auto library installation/loading for MeSH.XXX.eg.db. Note that this function is based on the NCBI Taxonomy ID embedded in METATDATA table in the sqlite3 file of LRBase.XXX.eg.db.
• The order of the parameters of cellCellSetting were changed as cellCellSetting(sce, lrbase, label, lr.evidence="all", color=NULL), and color parameter was changed as a optional parameter. If it is not specified, the color is automatically selected.
• The way to specify the celltype label was changed in cellCellSetting.
• The vignettes were modified.
• The convertNCBIGeneID is deprecated. The same functionality can be available as scTGIF::convertRowID instead.

scTensor 1.4.1-3

• A bug was fixed in .cellCellDecomp.Halpern()

scTensor 1.4.0

• verbose parameter was added in cellCellRank() and cellCellDecomp()
• set.seed(1234) was added in example and vignette
• .cellCellDecomp.CabelloAguilar() and .cellCellDecomp.Halpern() were added
• The num.iter 300　of NTD was changed to 30
• schex package was imported for visualing all the two dimensional gene plots
• The rule of cellCellRanks() based on the singular value of SVD was changed to a reconstruction error based rule, using NMF with matricised tensor in each mode
• The CCI-tensor is normalized in each frontral slice so that the total value is 1 (.frontal.normalization)

scTensor 1.2.1

• Some bugs were fixed

scTensor 1.2.0

• goenrich, meshenrich, reactomeenrich, doenrich, ncgenrich, and dgnenrich in cellCellReport() were added
• A bug related in sparse matrix in cellCellSetting() was fixed
• All the vignettes were updated
• A vignette for reanalysis of the results of scTensor was added
• Some bugs were fixed

scTensor 1.0.13

• Gene-gene network based on visNetwork was introduced

scTensor 1.0.11

• Some bugs were fixed

scTensor 1.0.10

• The tensor decomposition model was changed from non-negative Tucker3 to non-negative Tucker2
• assayNames option was added to specify which normalized assay data user want to use
• ftt option was removed (any normalization can be performed instead)

scTensor 1.0.7

• Some bugs were fixed

scTensor 1.0.6

• Replaced Dependency against biomaRt by AnnotationHub

scTensor 1.0.3

• Some bugs were fixed

scTensor 1.0.2

• Added upper in cellCellReport()
• Table grid lines in {ligand,receptor}*.Rmd were added
• Some bugs were fixed

scTensor 1.0.1

• Some bugs were fixed
• Three vignettes were added
• convertToNCBIGeneID() to ID conversion of input matrix

scTensor 1.0.0

• Package release

scTensor 0.99.22

• newCCSParams(), getParam(), setParam(), and cellCellSimulate() were added
• The hyperlinks to CMap (Connectivity Map) were embedded in the HTML report

scTensor 0.99.21

• Some bugs were fixed

scTensor 0.99.18

• Added Enrichment Analysis in cellCellReport()

scTensor 0.99.15

• Accepted in BioC 3.9

scTensor 0.99.6

• Revised and modified some parts

scTensor 0.99.0

• Package released