Version 2.1.4

The only change here is an attempt to fix the push to PyPI github action.

Version 2.1.3

Support for tensorflow 2.17 (thanks @jday1 and @ndalchau !)

Version 2.1.2

Support for tensorflow >= 2.16.1 (thanks @jday1 !)

Version 2.1.1

Minor bugfix release:

• Pin tensorflow version, as we do not support tensorflow 2.16.1 currently (#234)
• Fix tests so they run with pytest version 8

Version 2.1.0

This maintenance release fixes various minor issues. The models are not changed, and predictions should be identical to previous versions.

The main changes are:

1. Tensorflow dependency updated to 'tensorflow>=2.12.0'. We no longer run TF in version 1 compatibility mode.
2. Changed mhcflurry-predict-scan to be more intuitive. See #219
3. Make unit tests deterministic by correctly setting the random seed (we were setting it incorrectly before)
4. Use yaml.safe_load instead of yaml.load (fixes #215)
5. Switch to github actions for CI/CD instead of travis
6. Update Dockerfile to work with recent docker versions
7. Various small fixes to eliminate warnings

Thanks to @walid0925 for his helpful contributions on this release!

Version 2.0.6

Minor fixes

Version 2.0.4

Bugfix release

Fixes #195 and a few additional minor issues

pre-2.1

Not ready for use.

2.0.1

Fix the typo that caused #172

v2.0.0

Calling this version 2.0.0 to recognize that MHCflurry has evolved quite a bit in the last year. This release itself is not a major departure from the 1.6.0 version, however.

Version 2.0 introduces new models (BA/AP/PS). These follow the same overall design as in 1.6.0 but incorporate updated training data and some small tweaks:

• BA predictor: MS hits are assigned < 100 nM affinity instead of < 50 nM for training
• AP predictor: flanking window size decreased to 5 amino acids per flank from 15. Not seeing an accuracy decrease and it makes the models faster.

The code has been ported to tensorflow 2.0. Users will need to upgrade to a recent tensorflow to use MHCflurry.

Other changes:

• Support for percentile ranks for PS predictions
• Revamped Docker support
• Small fixes and docs updates