In this post London Calling ‘24 release we are delighted to introduce a broad collection of updates, improvements, and modifications to our EPI2ME Desktop Application, and the bioinformatics workflows that it can orchestrate. These updates were presented by Sirisha, Sarah, and Natalia during their presentations.

EPI2ME Desktop Application v5.1.14

Introduction of IGV to the Desktop Application - browse alignments and query your BAM files directly within the application - this is supported across all workflows that produce BAM format output.
Minimum workflow specifications - the Desktop Application will help you avoid running computionally and resource demanding workflows on inappropriate computer infrastructure.
Support for self signed certificates
Support for input data on non C: drives for Windows users
Various other general fixes and enhancements

IGV in EPI2ME Desktop — Figure 1. EPI2ME Desktop Application [v5.1.4] introduces the IGV browser as a plugin to further explore your mapped sequence data from workflows that produce BAM files. In this example a couple of SNPs identified by the wf-amplicon workflow are being investigated.

Workflow Updates

wf-template v5.1.4
- IGV support examples
- Sample grouping implementation
wf-basecalling v1.2.0
- dorado updated to 0.7.1
- New! Poly(A) tail length calling with the new parameter --poly-a-config (More info here: https://github.com/nanoporetech/dorado?tab=readme-ov-file#polya-tail-estimation)
- New! Demultiplexing support. The BC (barcode) tag will be added to the output CRAM file and demuxed BAMs will be produced in the folderdemuxed when you specify --barcode_kit. Demuxing parameters can be tuned with the addition of any of dorado demux args with the parameter --demux_args
wf-amplicon v1.1.0
- New! IGV support
- Some bug fixes and minor improvements to the report
wf-alignment v1.2.0
- New! IGV support
- Streamlined and simplified report; should use a lot less memory now
wf-metagenomics v2.10.0 wf-16s v1.2.0
- New! IGV support in the minimap2 approach
- Added:
  - Reads below percentages of identity (min_percent_identity) and the reference covered (min_ref_coverage) are considered as unclassified in the minimap2 approach.
  - The workflow now uses the `fastcat` read length and quality histograms instead of the per-read stats in the report process which reduces the memory use by the process.
- Fixed:
  - Files that are empty following the fastcat filtering are discarded from downstream analyses.
  - Checking the correspondence between the reference and ref2taxid now also works with compressed references.
  - “Can only use .dt accessor with datetimelike values” error in makeReport
  - “invalid literal for int() with base 10” error in makeRepory
  - Request less memory if kraken2_memory_mapping is used.
  - Show the percentage of each species when hovering over the taxonomy bar plot.
wf-clone-validation v1.3.0
- New! full_reference parameter for some additional assembly vs provided reference QC and a pLannotate bug fix
- BAM inputs now accepted (--bam)
wf-human-variation v2.2.3
- Improvements to memory management for alignment should finally rid users of a class of 137 errors.
- Minor performance improvements to the methylation subworkflow.
- Simplified phasing options.
- The workflow now supports ingress of multiple BAM files.
- Updates to Clair3, Longphase, Modkit and Spectre.
wf-aav v1.1.0
- Output BAM files tagged with assigned AAV genome type (AV:Z).
- BAM files can now be split by assigned genome type using the option output_genometype_bams.
- Medaka model is now automatically selected by identifying basecaller model with the input reads.
- New! Alignment results are now visible in the EPI2ME desktop app IGV viewer.
wf-single-cell v2.1.0
- This release Includes changes introduced since v2.0.0
- Several parts of the workflow are reworked to provide significant performance improvement
  - Users can expect ~ 3x speed up.
- Various updates to decrease peak memory usage including at the UMAP and expression matrix construction stages.
- BAM inputs now accepted (--bam)
- UMAP generation is now fast and so is now created by default (removed --plot_umaps)
- BAMs are output by sample not per chromosome (removed --merge_bam)
- Read batching is simplified with --adapter_scan_chunk_size and --process_chunk_size replaced with --fastq_chunk
- --kit_name and --kit_version replaced with --kit (eg: --kit 3prime:v3)
- 5prime:v2 kit added to available 10x kits
- Expression matrices are now output as sparse MEX files.
- updating Stringtie
- Several bug fixes including:
  - Report table displaying cell count off by one
  - Corrected number of unique genes and transcripts in report summary table
  - Fixed outputing incomplete BAM files introduced in v1.1.0
wf-somatic-variation v1.2.1
- General performance improvements of the workflow
- Introducing Severus as the somatic structural variant caller
- Updates to Clair3, modkit and snpEff
- Support for multiple BAM for single sample
wf-transcriptomes v1.2.0
- This workflow now accepts BAM input
- MA plot in output directory has been updated to match the MA plot in the report.

Please do let us know your feedback and we’re eager to hear wonderful new ideas for workflows and tutorials. Do watch our blog posts that will further describe some of these updates.