Major release

What's Changed

• Covalent docking, flex res specification in filelist, and bugfix for more than one ligand by @atillack in #171
• Analytical rotation coefficient gradient correction by @atillack in #184
• Add grid center output in DLG file Grid parameters section by @atillack in #189
• Fix for xml2dlg erroneously claiming no ligand or flexres was specified by @atillack in #195
• Flexres output for --gbest and --npdb by @atillack in #198
• Output interaction count for contact analysis even when no interaction was found by @atillack in #203
• Fix for potential buffer overflow in results processing by @atillack in #210
• Fixed pdbqt reading and switch sprintf's for snprintf's by @atillack in #212
• Added pose output reduction options by @atillack in #208
• Ensure later specified (i.e. on command line) modified pair parameters have priority by @atillack in #222
• Relative path for dpf import and batch files by @atillack in #225
• Increased maximum number of runs to 8192 by @atillack in #233
• Added output for grid map reading errors. by @atillack in #248
• Automatically dock entire directory contents by @atillack in #250
• Fix typo by @aaronegolden in #267
• Generate residue specific best coordinate files by @aaronegolden in #268
• Windows build and OpenCL/Cuda host code unification by @atillack in #255
• Automatically choose all available Cuda compute targets > 50 by @atillack in #270
• add meeko readthedocs to README docs section by @diogomart in #277
• Speeding up sum reductions in ADADELTA by using Tensor Cores by @L30nardoSV in #252

New Contributors

• @aaronegolden made their first contribution in #267

Full Changelog: v1.5.3...v1.6

Binaries

Attached are binaries for Linux, Windows, and macOS. For macOS, two sets of binaries are provided, "vanilla" macOS (not supporting OpenMP) and omp_macos binaries which requires libomp to be installed (i.e. using brew install llvm). The Linux version comes in OpenCL (x64 and aarch64) and Cuda 11 as well Cuda 12 versions (x64).