Change readme, news, and version (to 1.0.3)

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: NetCoMi
 Type: Package
 Title: Network Construction and Comparison for Microbiome Data
-Version: 1.0.2.9001
+Version: 1.0.3
 Author: Stefanie Peschel [aut, cre]
 Maintainer: Stefanie Peschel <[email protected]>
 Description: NetCoMi offers functions for constructing, analyzing, and comparing 

NEWS.Rmd

@@ -2,6 +2,49 @@
 output: github_document
 ---
 
+## NetCoMi 1.0.3 <img src="man/figures/NetCoMi_logo_800x400_300dpi.png" align="right" width="200" />
+
+This is a minor release with some bug fixes and changes in the documentation.
+
+### Bug fixes 
+
+* `netConstruct()` threw an error if the data had no row and/or column names, 
+which is fixed.
+
+* An edge list is added to the output of `netConstruct()` (issue [#41](https://github.com/stefpeschel/NetCoMi/issues/41)). See the 
+help page for details.
+
+* `SPRING`'s fast version of latent correlation computation (implemented in 
+[mixedCCA](https://github.com/irinagain/mixedCCA)) is currently not available 
+due to deprecation of the R package `chebpol`. The issue is fixed by setting 
+the `netConstruct()` parameter `measurePar$Rmethod` internally to "original" if 
+SPRING is used for association estimation.
+
+* In `plot.microNetProps()`: The `xpd` parameter is changed to `NA` so that 
+plotting outside the plot region is possible (useful for legends or additional
+text).
+
+* Labels in the network plot can now be suppressed by setting `labels = FALSE`
+(issue [#43](https://github.com/stefpeschel/NetCoMi/issues/43))
+
+* The `netCompare()` function threw an error if one of the permutation networks 
+was empty, i.e. had no edges with weight different from zero (issue [#38](https://github.com/stefpeschel/NetCoMi/issues/38)), 
+which is now fixed. 
+
+* Fix issues [#29](https://github.com/stefpeschel/NetCoMi/issues/29) and  [#40](https://github.com/stefpeschel/NetCoMi/issues/40), where permutation tests 
+did not terminate for small sample sizes. 
+Now, if the possible number of permutations (resulting from the sample 
+size) is smaller than that defined by the user, the function stops and returns 
+an error. 
+
+* Fix a bug in `diffnet()` (issue 
+[#51](https://github.com/stefpeschel/NetCoMi/issues/51)), where colors in 
+differential networks could not be changed.
+
+* `diffnet()` threw an error if the `netConstruct()` argument `jointPrepro` was 
+set to `TRUE`.
+
+
 ## NetCoMi 1.0.2
 
 This release includes a range of new features and fixes known bugs and issues.

NEWS.md

@@ -1,4 +1,53 @@
 
+## NetCoMi 1.0.3 <img src="man/figures/NetCoMi_logo_800x400_300dpi.png" align="right" width="200" />
+
+This is a minor release with some bug fixes and changes in the
+documentation.
+
+### Bug fixes
+
+-   `netConstruct()` threw an error if the data had no row and/or column
+    names, which is fixed.
+
+-   An edge list is added to the output of `netConstruct()` (issue
+    [\#41](https://github.com/stefpeschel/NetCoMi/issues/41)). See the
+    help page for details.
+
+-   `SPRING`’s fast version of latent correlation computation
+    (implemented in [mixedCCA](https://github.com/irinagain/mixedCCA))
+    is currently not available due to deprecation of the R package
+    `chebpol`. The issue is fixed by setting the `netConstruct()`
+    parameter `measurePar$Rmethod` internally to “original” if SPRING is
+    used for association estimation.
+
+-   In `plot.microNetProps()`: The `xpd` parameter is changed to `NA` so
+    that plotting outside the plot region is possible (useful for
+    legends or additional text).
+
+-   Labels in the network plot can now be suppressed by setting
+    `labels = FALSE` (issue
+    [\#43](https://github.com/stefpeschel/NetCoMi/issues/43))
+
+-   The `netCompare()` function threw an error if one of the permutation
+    networks was empty, i.e. had no edges with weight different from
+    zero (issue
+    [\#38](https://github.com/stefpeschel/NetCoMi/issues/38)), which is
+    now fixed.
+
+-   Fix issues [\#29](https://github.com/stefpeschel/NetCoMi/issues/29)
+    and [\#40](https://github.com/stefpeschel/NetCoMi/issues/40), where
+    permutation tests did not terminate for small sample sizes. Now, if
+    the possible number of permutations (resulting from the sample size)
+    is smaller than that defined by the user, the function stops and
+    returns an error.
+
+-   Fix a bug in `diffnet()` (issue
+    [\#51](https://github.com/stefpeschel/NetCoMi/issues/51)), where
+    colors in differential networks could not be changed.
+
+-   `diffnet()` threw an error if the `netConstruct()` argument
+    `jointPrepro` was set to `TRUE`.
+
 ## NetCoMi 1.0.2
 
 This release includes a range of new features and fixes known bugs and

readme.Rmd

@@ -103,7 +103,7 @@ Microbial associations are compared between the two experimantal settings
 ## Overview of methods included in NetCoMi
 
 Here is an overview of methods available for network construction, together with
-some information on the implementation in R:
+some information on their implementation in R:
 
 **Association measures:**
 
@@ -156,20 +156,33 @@ TSS, CSS, COM, VST, and the clr transformation are described in [Badri et al., 2
 ## Installation
 
 ```{r install, eval=FALSE}
-#install.packages("devtools")
+# Required packages
+install.packages("devtools")
+install.packages("BiocManager")
 
-devtools::install_github("stefpeschel/NetCoMi", dependencies = TRUE,
+# Install NetCoMi
+devtools::install_github("stefpeschel/NetCoMi", 
+                         dependencies = c("Depends", "Imports", "LinkingTo"),
                          repos = c("https://cloud.r-project.org/",
                                    BiocManager::repositories()))
 ```
 
 If there are any errors during installation, please install the missing 
 dependencies manually.
 
-Packages that are optionally required in certain settings are not installed 
-together with NetCoMi. These can be automatically installed using:
+In particular the automatic installation of 
+[`SPRING`](https://github.com/GraceYoon/SPRING) and  [`SpiecEasi`](https://github.com/zdk123/SpiecEasi) (only available on 
+GitHub) does sometimes not work. These packages can be installed as follows:
 
 ```{r install2, eval=FALSE}
+devtools::install_github("GraceYoon/SPRING")
+devtools::install_github("zdk123/SpiecEasi")
+```
+
+Packages that are optionally required in certain settings are not installed 
+together with NetCoMi. These can be installed automatically using:
+
+```{r install3, eval=FALSE}
 installNetCoMiPacks()
 
 # Please check:
@@ -211,7 +224,7 @@ The data are filtered within `netConstruct()` as follows:
 
 - Only samples with a total number of reads of at least 1000 are included 
 (argument `filtSamp`). 
-- Only the 100 taxa with highest frequency are included (argument `filtTax`).
+- Only the 50 taxa with highest frequency are included (argument `filtTax`).
 
 `measure` defines the association or dissimilarity measure, which is `"spring"` 
 in our case. Additional arguments are passed to `SPRING()` via `measurePar`. 
@@ -232,10 +245,10 @@ weights in the network plot.
 The `verbose` argument is set to 3 so that all messages generated by 
 `netConstruct()` as well as messages of external functions are printed.
 
-```{r single_spring}
+```{r single_spring, eval=FALSE}
 net_single <- netConstruct(amgut1.filt,
                            filtTax = "highestFreq",
-                           filtTaxPar = list(highestFreq = 100),
+                           filtTaxPar = list(highestFreq = 50),
                            filtSamp = "totalReads",
                            filtSampPar = list(totalReads = 1000),
                            measure = "spring",
@@ -249,6 +262,21 @@ net_single <- netConstruct(amgut1.filt,
                            seed = 123456)
 ```
 
+```{r single_spring2, eval=FALSE, echo=FALSE}
+saveRDS(net_single, file = "./data/net_single_spring.rds")
+```
+
+```{r single_spring3, eval=TRUE, echo=FALSE}
+net_single <- readRDS("./data/net_single_spring.rds")
+```
+
+
+**Note:** Since the fast version of latent correlation computation (for which 
+the [mixedCCA](https://github.com/irinagain/mixedCCA) package is used in `SPRING`) 
+is currently unavailable, the `SPRING` parameter `Rmethod` is internally set to 
+"original". Accordingly, execution time is currently considerably increased for 
+the SPRING approach.
+
 **Analyzing the constructed network** 
 
 NetCoMi's `netAnalyze()` function is used for analyzing the constructed 
@@ -587,13 +615,15 @@ The two resulting phyloseq objects (we ignore the group
 
 We select the 50 nodes with highest variance to get smaller networks.
 
-```{r netcomp_spring_1}
+```{r netcomp_spring_0}
 # devtools::install_github("vmikk/metagMisc")
 
 # Split the phyloseq object into two groups
 amgut_split <- metagMisc::phyloseq_sep_variable(amgut2.filt.phy, 
                                                 "SEASONAL_ALLERGIES")
+```
 
+```{r netcomp_spring_1, eval = FALSE}
 # Network construction
 net_season <- netConstruct(data = amgut_split$no, 
                            data2 = amgut_split$yes,  
@@ -610,6 +640,14 @@ net_season <- netConstruct(data = amgut_split$no,
                            seed = 123456)
 ```
 
+```{r netcomp_spring_1a, eval=FALSE, echo=FALSE}
+saveRDS(net_season, file = "./data/net_season.rds")
+```
+
+```{r netcomp_spring_1s, eval=TRUE, echo=FALSE}
+net_season <- readRDS("./data/net_season.rds")
+```
+
 Alternatively, a group vector could be passed to `group`, according to which
 the data set is split into two groups: