2.2.8

Functions

amp_subset_taxa

• Added remove = TRUE/FALSE to indicate whether the taxa in tax_vector are the ones being removed from the data or the ones being kept

amp_alphadiv

• It's now possible to rarefy when rarefy < minreads

amp_ordinate

• Added unifrac support, weighted and unweighted
• Added status messages when using distance-based ordination on huge data

amp_timeseries

Has been overhauled

• If the data contains more than one sample from the same date, then the average is calculated unless the samples are split by a group variable defined by group_by
• group_by not works correctly
• Added plotly support

Data

• Added UPGMA trees in example data
• Updated the taxonomy to the newest MiDAS taxonomy

2.2.7

amp_load

• Added a "tree" element in ampvis2 objects for phylogenetic trees to be loaded with ape::read.tree() (for future new features)
• When no metadata is provided (not recommended), dummy metadata is now created

amp_subset_samples

• Added removeAbsent argument so user decides whether OTU's that may be absent in all samples after a subset are removed or not
• Bug fixes:
  • Setting minreads and normalise = TRUE together resulted in incorrect subsets, and even returned empty data when minreads set to 100 or more.

amp_alphadiv

• Rarefy set to a lower value than the smallest amount of reads in any sample now throws a warning instead of an error
• Bug fixes:
  • Fixed an issue where the results were sometimes NA
  • Fixed an issue where the order of the results per sample were scrambled in some cases

amp_heatmap

• Facet strip text size increased to 10

amp_rarecurve

• Bug fixes:
  • Fixed an issue where an error occured when a column was named "SampleID", which wasn't the first column

2.2.6

Webpage

• Various webpage updates, including a new FAQ section.

ampvis2 functions

amp_ordinate

• Filtering is based on percentages, but the data used for ordination is untouched. Previously the data was just normalised at the start.
• sample_colorframe: Can now take a variable in the metadata, so groups within groups can be framed.

amp_heatmap

• Has been thoroughly tested and confirmed to show correct values. The mean group abundances calculated by ampvis1 heatmaps cannot be trusted, especially when there are many 0's in the group.
• Removed detailed_output = TRUE/FALSE:
  • The data used for ggplot's is already inherited in the returned ggplot2 object as $data.
• Added textmap = TRUE/FALSE: prints a text version of the heatmap in the console to extract the data in the heatmap.

amp_alphadiv

• Has been completely rewritten, but with the same features.
• Added an ObservedOTUs column with the number of unique OTUs per sample.
• Added a richness = TRUE/FALSE argument which calculates Chao1 and ACE richness estimates per sample.
• More error messages and checks

amp_load

• ape::read.dna to load FASTA files has been replaced with ape::read.FASTA:
  • This fixed an issue where the OTU names didn't load when the otutable was generated with different linux scripts, which caused errors.

amp_rarecurve

• Throws an error when stepsize is higher than max#reads.

2.2.5

Minor fixes:

• Adjusted onAttach message.
• When there are excess sample ID's in either otutable or metadata, the taxonomy table is now also subsetted and OTUs with 0 read abundance in all samples are silently
• amp_rename now removes any number of underscores in taxa names.
  • This resulted in incorrect taxa names and therefore a seemingly higher diversity when they contained different number of underscores even though the same name. Fx g____Tetrasphaera would result in a Genus called __Tetrasphaera, not just Tetrasphaera.

2.2.4

• amp_load now handles single sample data correctly
• fixed filter_species in amp_ordinate (added percentages again, brainfart)

2.2.3

• library(ampvis2) now checks for new github release versions and notifies the user if a new version is available. This does NOT involve the master branch, only releases.
• amp_ordinate():
  • transform="sqrt" wasn't working with nmds/pcoa
  • Adjusted warning messages
  • Added sample_point_size to adjust the sizes of the sample points
  • Removed part of the code that transformed to percentages before ordination. So ordination now more precise, but maybe a bit slower (1 sec)

v2.2

Numerous bug fixes and improvements compared to 2.1. Specifically amp_load has been improved to be more robust and now handles otutables loaded in different ways, with more meaningful error messages.

Revised data structure and more complete documentation

2.1

plotly was missing in imports!

For backwards compatibility

2.0

Minor fix to text positions in heatmaps.