[Segway-users] Segway 3.0 released

Roberts, Eric Eric.Roberts at uhnresearch.ca
Tue Feb 11 08:16:46 PST 2020


We are pleased to announce the availability of Segway 3.0!

This release is focused on two new major features of Segway.

First, we have added a more modular interface to Segway. Specifically for each Segway task, such as train and annotate, we have added the steps "init", "run", and "finish", and specific to train, "run-round". Note that these are optional, and simply specifying the task without any subtask will run as it did before. However, as a result of these changes, there's a minor breaking change in interface. Options (which are mostly technical, such as '--cluster-opt') that are shared across all tasks must be specified before any task, and task-specific options must be specified after the task. The new Segway commands have the following layout: "segway [global_args] COMMAND [args]...". If you did not use any global arguments, your commands will be identical. Upgrading is recommended and moving the options before the task as necessary is not anticipated to be a difficult fix for users. A paper discussing applications of the modular interface is planned for a later date.

Second, we have added the ability to add virtual evidence to the Segway model. Effectively this allows you to specify a prior probability for any label of your choosing at any position in your genome. This a form of semi-supervised learning and further results and discussion can be found in a preprint by Rachel Chan on Biorxiv (https://biorxiv.org/cgi/content/short/2020.0130.926923v1). The paper uses Segway to model transcription start site prediction and compares using virtual evidence to using the traditional unsupervised approach. For more details on how to use virtual evidence in Segway see the documentation (https://segway.hoffmanlab.org/doc/3.0/segway.html#virtual-evidence).

Segway (and Genomedata) are both available on Bioconda. With Bioconda installed, to install segway run 'conda install segway'. Segway can still be installed through older methods using "pip", see the Quickstart Guide (https://segway.readthedocs.io/en/latest/quick.html) for details.

Here is a full list of changes made since the last announcement here:

Enhancements:
* segway: added modular interface for each task. subtasks include init, run, finish,
and run-round for training specifically
* segway: added cluster-specific max job polling times (thanks to Rachel Chan
for diagnosis)
* segway: added virtual evidence as an option to the Segway model

This e-mail may contain confidential and/or privileged information for the sole use of the intended recipient.
Any review or distribution by anyone other than the person for whom it was originally intended is strictly prohibited.
If you have received this e-mail in error, please contact the sender and delete all copies.
Opinions, conclusions or other information contained in this e-mail may not be that of the organization.

If you feel you have received an email from UHN of a commercial nature and would like to be removed from the sender's mailing list please do one of the following:
(1) Follow any unsubscribe process the sender has included in their email
(2) Where no unsubscribe process has been included, reply to the sender and type "unsubscribe" in the subject line. If you require additional information please go to our UHN Newsletters and Mailing Lists page.
Please note that we are unable to automatically unsubscribe individuals from all UHN mailing lists.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman13.u.washington.edu/pipermail/segway-users/attachments/20200211/b66c7868/attachment.html>


More information about the Segway-users mailing list