- model.write_data() now writes chain IDs to the files for 'NGH'
and 'SSM' output.
- model.read() and model.write() now understand mmCIF format
author-provided residue numbers (these are always used by MODELLER if
available, rather than the canonical seq_id, which always counts from 1).
- model.write() now uses hybrid-36 notation in PDB files for atom serial
numbers >= 100000.
- Bugfix: correctly handle alignment expansion after truncation. In some
rare cases, lengthening the alignment with alignment.append() or
alignment.append_sequence() would add spurious repeated sequences to the
alignment. This should now be handled correctly.
- Bugfix: don't exhaust stack on Windows machines when processing large
structures with model.write_data().
Automatic builds
2019-06-19
