Mean of top 20 / bottom 10 addin: - New addin helps to evaluate whether the staining quality is likely to produce good unmixing. See https://akoyabio.github.io/phenoptrReports/articles/top_20_bottom_10_report.html for details.
Staining consistency report: - New addin creates a report showing mean expression of a selected marker across multiple samples.
Analysis addin: - Fix problem computing mean expression for phenotypes with formulas in their definition #47 - Allow selecting (and reporting on) multiple score summary files #49 - Tabs with H-Score results include the scored marker name (in support of #49)
Spatial map viewer: - Add support for selecting the desired view when the exported composite files contain multiple views.
Consolidate and summarize addin: - Add a dummy
Phenotype column if none is present #48
Merge addin: - Read files to merge into memory and write once, instead of appending
Component levels report: - Works correctly when “Export data from pairs plots” is checked #50
Analysis addin: - Fix problems with chart generation #46
Consolidate and summarize addin: - Add support for reading (and saving) only fields needed in phenoptrReports.
Analysis addin: - Preserve (or recreate) TMA information in the analysis output files. - Only read cell seg fields needed for the analysis. - Fix network file paths in
Component levels report: - Fix histograms to work with more than 11 components. - Omit pairs plots with more than 10 components. - Add correlation heatmaps for each file
Misc: - Update to work with multi-schema phenotypes - Fix report generation to work when paths contain spaces - Require R 4.0.0 or higher for compatibility with latest rtree
Consolidate and summarize addin: - Reduce memory use to allow consolidation of larger files
Bug fixes: - Nearest neighbor and count within summaries work correctly if detailed data is not requested. - Correctly double-escape all backslashes in the slide id prefix. - Improved error checks and error messages
Analysis addin: - Nearest neighbors and count within are computed per tissue category and for all cells in any included category (#39). - Include expression-based phenotype columns and measured expression columns in count_within.txt and nearest_neighbors.txt (#36). - Output has nicer names for expression-based phenotypes (#36).
Spatial map viewer addin: - Fix “Save all” to work on a Mac
Analysis addin: - Remove summary lines from the Count Within worksheet, they are not correct (#35)!
Spatial map viewer addin: - For the usual use where the client browser is running on the user’s computer, the “Save all” saves directly to the local file system without creating an intermediate zip file. This is faster and avoids transfer of a potentially large zip file.
Misc: - Fix to work correctly with dplyr 1.0.0.
Analysis addin: - The Count Within worksheet includes summary counts across all tissue categories when multiple categories are reported. - Fix chart report to run when “Total Cells” is the only phenotype defined (#32). - Fix chart report to handle “All Cells” correctly. - Fix H-Score report to always include every combination of Slide ID (or Sample Name) and Tissue Category. - When computing mean expression for All tissue categories, don’t propagate NA. The All numbers reflect the valid tissue categories. - Fix to work with latest
Spatial map viewer addin: - Don’t require Slide ID in the source cell data file.
Analysis addin: - Don’t require tissue segmentation in the inForm project (#4). - Add the ability to compute H-Score for individual phenotypes (#30). When selected, additional H-Score worksheets and charts will be created in the output reports. - When “Use regular expressions” is checked, the regular expression will match anywhere in the Slide ID, not just the prefix. - Fill in missing tissue categories so all slides have all categories reported (#1). - Report missing / invalid values consistently as #N/A, never as #NUM.
Merge addin: - Fix a problem that caused data corruption when cell seg data is formatted with comma as decimal separator (#31).
Misc: - Use the RStudio folder chooser when available, rather than the Windows native chooser. The RStudio chooser is nicer and gives access to network drives (#29).
Analysis addin: - Reverse the y-axis of heatmaps so that the identity runs from the top left to bottom right.
Spatial map viewer improvements: - Add “touching cells” visualization and related data. - When saving plots, the viewer optionally saves the associated data (#22). - Spatial map viewer works correctly with inForm data containing field coordinates (vs slide coordinates) (#16).
Consolidate and summarize addin: - Add a Slide ID column with value “None” if none is present.
Component levels report: - Optionally export data from pairs plots as TSV files.
Misc: - Now requires tidyr >= 1.0.0.
Spatial map viewer improvements: - Dramatic improvement in time to render the charts, especially with 2x2 fields.
Analysis addin: - Save nearest neighbor and count within detail as tab-separated text files instead of CSV files. This allows the detail files to be used as input to a second run of the analysis addin. - Improved heatmaps in Charts output: - Maximum of eight heatmaps per page - Split the facet titles so the full Slide or Annotation ID is visible - Better
nearest_neighbor_summary - Avoid warnings from
max and return
NA for all values when there are no examples of a phenotype (#20).
Unmixing quality report: - Tables show components as ordered in the data, not in alphabetical order. - Pixel intensity chart uses actual pixel intensity as labels on the x-axis, rather than log10 values.
Component levels report: - Charts use actual pixel intensity as labels on the x-axis, rather than log10 values.
Spatial map viewer improvements: - Calculates nearest neighbors on the fly - Works with data from
Consolidated_data.txt (doesn’t need
nearest_neighbors.csv) - Supports multiple positivity and expressions in phenotype selectors!
Component levels report: - Allow the user to enter the quantile(s) to display on the histograms. If two or more quantiles are entered, include a signal-to-noise table in the report, showing the ratio of the highest quantile to the lowest.
Bug fixes: - Fixed a problem that caused the analysis addin to incorrectly report zero cells in a phenotype (#15).
Spatial map viewer addin: - New addin provides a field-based viewer of nearest neighbor relationships. See the Visualizing spatial relationships tutorial for details.
Analysis addin: - Phenotype definitions may include valid expressions such as
~`Membrane PDL1 (Opal 520) Mean`>5.
Bug fixes: - Fix error during calculation of N/A and Total cell in cell seg summary report
Bug fixes: - Fix calculation of N/A and Total cell count in cell seg summary report (#13)
Analysis addin: - Add support for aggregation by a selectable column (#3). This allows reporting by Sample Name or Annotation ID as well as by Slide ID. - Detailed output of nearest neighbors includes the Cell IDs of nearest neighbors as provided by
phenoptr::find_nearest_neighbors. - Recognize and correctly read inForm data which uses comma as the decimal separator. (Requires phenoptr >= 0.1.6.9000; related to akoyabio/phenoptr#8). - More robust
find_common_prefix function works with empty and numeric Slide IDs.
Bug fixes: - Fix the way the analysis addin handles special characters in Slide ID (#9). - Fix file chooser to work on RStudio Server.
Analysis addin - Optionally save nearest-neighbor and count-within details for each cell. - Reverse the fill scale of the nearest neighbor heatmaps so red represents closer cells. - Write session info to
session_info.txt in the output directory.
Unmixing quality report - Add “Guidance” section.
Component levels report - Show a single quantile (99.9 percentile) in the signal histograms. - Omit Autofluorescence from the pairs plots. - Use a common scale on the individual plots within each pairs plot to avoid exaggerating small signals. - Add “Guidance” section. - Remove “Signal to noise” section.
Merge addin - More forgiving handling of duplicated records. It now ignores the duplicates with a warning. - Less noisy console output.
New features: - Analysis app optionally creates tabular summaries and heatmap visualizations of nearest neighbor distance and count within radius for all phenotype pairs. - Component levels report includes pair plots of component vs component for each source image.
Bug fixes: - Component levels report - Write CSV files to the export directory, not its parent. - Fix height of “Pixel intensity by component” charts. - Consolidate merged files - Fix to recognize “Annotation ID” as the field name. - Analyze consolidated data - Update for dplyr 0.8 to fix #N/A values in Total lines of Cell Counts, Cell Percents and Cell Density sheets - Clean up chart output - “All” tissue category is last in charts which have it, same as “Total”. - Fix “Cell Count” axis label - Fix legend in H-Score chart - “name repair” in readxl version 1.2.0 created unexpected column names which were shown in the legend.
New features: - The “Merge cell seg files” addin now supports searching sub-directories for target files. - The “Consolidate and summarize” addin remembers the last directory for subsequent selections. - More informative error message if consolidation fails due to mis-matched data files.
Bug fix: -
compute_density works with “square microns” spelled as inForm does.
New features: - Component levels report shows signal levels of bright and dark pixels in multiplex images.
Misc: - Unmixing quality report works when DAPI and AF components are not present. -
consolidate_and_summarize_cell_seg_data processes files pairwise for reduced memory requirements. - Better error handling in
consolidate_and_summarize_cell_seg_data. - Removed
New features: - New RStudio Addin - “Merge cell seg files” merges inForm output from individual fields, similar to the inForm Merge tab. -
merge_cell_seg_files function performs the merge.
Misc: - Update the link to Phenoptics home page. -
compute_mean_expression_many ignores duplicate parameter definitions instead of stopping with an obscure error. -
compute_density_from_cell_summary doesn’t require a
Phenotype column in the cell seg summary data. -
split_phenotypes doesn’t require a
Confidence column in the source data. -
split_phenotypes recognizes whitespace as a phenotype separator.
New features: - New unmixing quality report. - Mean expression sheet includes tissue category “All” with mean expression across all requested categories. - Chart report includes count and density charts with truncated Y-axis if any count or density > 2000.
New features: - Add version number, date stamp and documentation link to generated script file #2 - Add H-Score chart to the generated charts - Add Akoya logo to footer in chart report
Bug fixes: - Fix order of slides in Mean Expression report - Don’t crash cell seg summary report when a slide has only one phenotype - Correctly report number of fields for data derived from a whole-slide scan