-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathCHANGELOG
100 lines (85 loc) · 3.34 KB
/
CHANGELOG
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
v1.5.0
01/24/2025
NEW FEATURES
- AlphaFold3 updated with v3.0.1
v1.4.0
01/13/2025
NEW FEATURES
- AlphaFold3
- Bootstrap using Docker containers on Docker Hub 4geniac
BUGFIXES
- Remove double-quotes for boolean parameter in json files
- Numpy forced to version 1.26.4 in nanoBERT container
DOCUMENTATION
- Detail outputs and explain how the different annotations must be installed
- Examples of different channels patterns explained in the nextflow code
v1.3.0
11/20/2024
NEW FEATURES
- AFMassive:
* All the models are computed in parallel
* A process has been added to compute relaxation
- AlphaFill:
* tool to predict the binding of missing compounds to predicted 3D protein structure
* multiqc report with the predictions
* take into account that the json output can include different value for the same transplant, or some transplant may have missing information
- DynamicBind:
* Add test data
- New tools:
* nanoBERT: predict the nanobody mutational space
* DiffDock: molecular docking
- Result folder:
* organize publishDir for AlphaFold and AfMassive in order to avoid three folder levels.
- Structure prediction with multimer:
* compute metrics pDockQ and iPAE
- Process:
* Generate PAE matrix in JSON format for the best model
- multiqc reports:
* legends and references have been added in the reports
* proteinStruct includes the ranking_debug.json table and multimer metrics (when available) for all models
* new option to generate reports from existing results:
--htmlProteinStruct
* PAE plots: max value set to 31.75
* fasta sequence is displayed at the beginning of the report
* Mosaic plot with all 3D structures
* In the plots, model name legend which contains rank start from 0 (instead of 1) to be consistent with the naming produced by AlphaFold
- code refactoring:
* implementation of nextflow subworkflows
* ensure that the pipeline can resume
* adaptive memory for metricsMultimer process
* pylint / shellcheck scripts, rename in-house script with ap_ prefix
* set PATH in container recipes
* process and files related to multiqc hhave been renamed with mqc
v1.2.0
03/08/2024
NEW FEATURES
- AFMassive:
* massiveFold has been renamed by AFMassive
* --dropout set as default in --afMassiveOptions
- AlphaFold:
* memory has been increased for both alphaFoldSearch and alphaFold processes
- ColabFold:
* use --save-all option to save pickle files
* --colabSearchHelp option
* add amber relaxation by default
- DynamicBind:
* tool for molecular docking
- multiqc reports
* plots
* print software version from containers (get_version.sh)
- Groovy functions moved from main.nf to lib/function.nf
- add stub section in process for dry-run with -stub-run
BUG FIXES
- change regex to get protein name from fasta filename
v1.1.0
02/22/2024
NEW FEATURES
- MassiveFold updated with AFmassive v1.1.0
- MassiveFold & AlphaFold: msas performed in parallel for each chain with the multimer mode
- option --onlyMsas to perform only multiple sequence alignments
- option --fromMsas to perform 3D structure prediction from existing multiple sequence alignments
- msas test data have been added
v1.0.0
02/02/2024
NEW FEATURES
- Prediction of protein 3D structures with AlphaFold, MassiveFold and ColabFold