FAQ
The following are some frequently asked questions from our users. For information to help troubleshoot common problems see our troubleshooting guide. You can also use the site search in the top menu to find information on specific topics, including searching these FAQs.
EPI2ME Desktop Application
How do I check the version of EPI2ME Desktop I have installed? ⧉
Click on the settings cog in the left panel.
Does EPI2ME Desktop need an internet connection? ⧉
To deliver the best possible experience in delivering up-to-date bioinformatic analyses, EPI2ME Desktop requires an internet connection to download workflows the first time they are run. After workflows are downloaded, an internet connection is not typically required for further use. Some connectivity may be required for downloading biological reference data from, for example, NCBI.
What internet resources does EPI2ME Desktop need to access? ⧉
The application will attempt connections to the following domains: epi2me.nanoporetech.com, api.epi2mecloud.com, amazonaws.com, api.github.com, docker.io, docker.com, oxfordnanoportal.com, heapanalytics.com, amazonaws.com, okta.com, oktacdn.co
What IP range do I need to allow for cloud analyses to work? ⧉
While workflows can be run offline, some online resources might be required at setup. The AWS documentation online can give you the IP addresses for the AWS services in question. You can find them here.
Do I need to install Java and Nextflow in order to use EPI2ME Desktop? ⧉
This is not necessary, the application manages the install of these software items if required. See our installation guide for more information.
Do I need to install Docker Desktop in order to use EPI2ME Desktop? ⧉
This depends on your operating system. Docker Desktop is not required on Windows, but for MacOS you should ensure it is installed. For Linux computers, Docker Engine is advisable. See our installation guide for more information.
How do I update the application? ⧉
On starting the application users will be informed if an update is available for the application. Updates to the application can also be downloaded from the downloads page.
Can I use EPI2ME Desktop on my computing cluster? ⧉
Our Nextflow workflows will work natively with most cluster systems. Refer to the Nextflow documentation and speak to your system administrators. EPI2ME Desktop can be made to work with a cluster through its generic Nextflow config handling, but this is not directly supported. Please contact us.
Can I install EPI2ME Desktop on my GridION or PromethION? ⧉
You may install and run EPI2ME Desktop on your device. Follow the Linux (Debian) installation instructions. EPI2ME Desktop will be included in future versions of the device software.
Warning
On GridION and PromethION sequencing devices, it is advised to not run data analysis through EPI2ME Desktop whilst sequencing is in progress.
How long do I have to download my cloud analysis results? ⧉
The results of cloud analyses are kept for 2 weeks from the completion of the analysis. Results will be automatically downloaded as they are created for any analyses started since the app was last started.
How do I find my Cloud Analysis ID? ⧉
Select your EPI2ME Cloud analysis. Click options in the top right. Select report issue, your Cloud Analysis ID will be available on screen
Which workflows can I run in EPI2ME Cloud? ⧉
At present, only Oxford Nanopore Technologies authored workflows can be run in the EPI2ME Cloud platform. EPI2ME does not currently permit arbitrary code to be run within the platform; externally written workflows are not supported.
Running And Managing Workflows
Can I run workflows on my GridION or PromethION? ⧉
You may install and run EPI2ME Desktop on your device. Follow the Linux (Debian) installation instructions. EPI2ME Desktop will be included in future versions of the device software. Nextflow should not be run on from a command prompt on devices - the preinstalled version of Nextflow is intended for the sole use of MinKNOW.
Warning
On GridION and PromethION sequencing devices, it is advised to not run data analysis through EPI2ME Desktop whilst sequencing is in progress.
Can I use data from other technologies? ⧉
EPI2ME workflows are designed and optimised to work with Oxford Nanopore Technologies’ sequencing data. They will typically not produce optimal results with data from other sequencing technologies.
Can I go back to a previously submitted run to edit the parameters and rerun the workflow, or save a set of parameters to load into a new run? ⧉
This is not currently possible in EPI2ME Desktop.
Can I run more than one workflow at a time? ⧉
Multiple workflows can in principle be run simultaneously. EPI2ME Desktop does not however monitor the total resource used by workflows it starts to ensure the system is not put under undue resource pressure: users should be aware that running many workflows simultaneously may affect system performance and stability. You can watch the status of running workflows by clicking the instances clock icon.
Can I run workflows offline? ⧉
You can run workflows offline. You will need to run and install a workflow once whilst online and then the workflow should run fine offline once it has done initial set up.
Do the workflow minimum CPU requirements refer to threads or cores? ⧉
The minimum CPU requirements strictly refers to the number of threads; it is threads (logical processors) that are used in the resource accounting.
However some workflows will not scale efficiently when simultaneous multithreading (hyperthreading) is enabled.
For example users may find that a workflow with an 4 core, 8 thread processor runs a workflow no faster when set to use 8 threads than when set to use 4.
Can HTML reports be viewed offline? ⧉
Yes, our HTML reports are fully standalone and do not require an internet connection to view. You can open the HTML report in any web browser, even if you are offline.
Where can I find a list of the expected output for each workflow? ⧉
There is a list of expected outputs for each workflow in the GitHub repository for each workflow, scroll down in the README to outputs.
How do I find the output location from running a workflow? ⧉
Click the open instance button in the workflow overview tab.
Where can I find example data files to test the workflows on my computer? ⧉
The workflow shows in EPI2ME Desktop as completed, but there was no output. What steps should I take next? ⧉
Click open instance, and see if there are any output files present. If not then it is likely an error with the platform. Please contact us.
The workflow I ran ended with an error, what steps should I take next? Which files are needed to send in with a support ticket? ⧉
Consult our guide on interpreting Nextflow errors and workflow exit codes which may help you diagnose the problem with your workflow. If the workflow fails again please contact us
I would like to clear more disk space after running workflows, which files are safe to delete? ⧉
You can manually delete the work folder of a run, click open instance from the workflow overview page and manually delete the work folder. If you are completely finished with a workflow and no longer need any of the data you can delete the entire directory.
Do EPI2ME Labs workflows support barcoded data? ⧉
Most of our workflows do support barcoded data but not all. If a workflow has a sample sheet parameter, then it does support barcoded data.
To run with multiple barcodes, you should input a directory that contains subdirectories one for each barcode. Optionally a sample sheet can be included to replace barcodes with sample names in any output files.
eg. input directory structure
Sample sheet files should be laid out according to the MinKNOW sample sheet specification. For EPI2ME Labs workflows a file minimally needs the `barcode` and `sample_id` fields:
├── barcode01
│ ├── reads0.fastq
│ ├── reads1.fastq
│ └── reads2.fastq
├── barcode02
│ ├── reads0.fastq
│ ├── reads1.fastq
│ └── reads2.fastq
└── barcode03
└── reads0.fastq
barcode,sample_id
barcode01,sample01
barcode02,sample02
barcode03,sample03
My system has sufficient memory, but the workflow exits with an Out of Memory error ⧉
Even if your system exceeds the minimum memory requirements for a given workflow, your operating system also requires some of that memory. For example if the minimum requirement for the workflow is 31GB, and your system has 32GB RAM, the OS will likely require more than the 1GB difference.
More about Workflows
Where can I find more information about workflows and what they does? ⧉
In the workflows page contains all documentation for each of our workflows. The Pipeline Overview section which explains how your data is processed and which are the steps and tools that are used. This information is also accessible when installing a workflow in the EPI2ME Desktop Application.
Which file on a workflows' GitHub page provides a list of the tools used, their versions, and parameters to run the command? ⧉
Software versions used are listed in the reports generated by EPI2ME Labs workflows. The parameters used can be found by inspecting the workflow source code.
Where can I find the code and commands used by the workflows? ⧉
Why do EPI2ME workflow reports show a different quality score distribution to the one shown in MinKNOW? ⧉
The dorado basecaller, and so MinKNOW, calculates a read’s average quality score excluding 60 bases from the ends of the read. Those excluded per-base quality scores are typically lower that those in the central portion of the read.
The fastcat tool used in the data ingress stage of all our workflows is responsible for building quality score histograms. It does one of two things to collate this information:
- if it can find a meta data tag in the input file for the basecallers mean qscore it uses that, or
- it recomputes the mean using ALL of the bases present.
In the first case, the EPI2ME workflow reports should show quality score histogram’s that appear similar to those from MinKNOW. In the second case, results may differ. fastcat uses all the bases for two reasons:
- this is what was done historically, and
- it does not know whether the basecaller trimmed the read or not: if it excluded some quality scores from the calculation, it could end up overcompensating.
There is unfortunately no universally correct solution that fastcat can apply when recomputing the quality score.