DeltaGroupNJUPT / Vina-GPU-2.0

Vina-GPU 2.0 accelerates AutoDock Vina and its related commonly derived docking methods, such as QuickVina 2 and QuickVina-W with GPUs.
Apache License 2.0
89 stars 29 forks source link

Bad guys hiding in massive files cause errors. #22

Open xukm3 opened 1 year ago

xukm3 commented 1 year ago

我们从VirtualFlow官网(https://virtual-flow.org/real-library?check_logged_in=1)下载了约100万个pdbq格式的化合物,打算利用win vina-GPU进行虚拟筛选。但是下载的数据库存在一些问题,比如有些pdbqt文件用pymol打开显示出某些原子未成键(见图 image ) ),有些文件包含Si原子因此被识别为非法的pdbqt文件( image

每一次运行vina-GPU时,如果输入的配体文件中包含一个或多个有问题的文件,则不能正确运行,但是大量的化合物无法进行费时费力的排查并逐一剔除问题文件。我们设想,在运行vina-GPU时是否是一次读取文件夹中的所有pdbqt格式配体然后进行并行对接,如果是这样,在遇到某些有问题的文件时能否有办法跳过或忽略它,从而避免报错。请问您能否给我们宝贵的意见,谢谢。

We downloaded about 1 million compounds in pdbq format from VirtualFlow (https://virtual-flow.org/real-library?check_logged_in=1) to use win Ina -GPU for virtual filtering. However, there are some problems with the downloaded database, such as some pdbqt files opened with pymol showing that some atoms are not bonded (see Figure(image), Some files contain Si atoms and are therefore identified as illegal pdbqt files ( image ).

Every time you run vina-GPU, if the input ligand file contains one or more problematic files, it will not run correctly, but a large number of compounds cannot be time-consuming and laborious to check and remove the problem files one by one. We wonder if, when running vina-GPU, all the pdbqt format ligands in the folder are read at once and then interfaced in parallel. If so, if there is a way to skip or ignore some problematic files, so as to avoid errors. Could you please give us your valuable advice? Thank you.

Every time you run vina-GPU, if the input ligand file contains one or more problematic files, it will not run correctly, but a large number of compounds cannot be time-consuming and laborious to check and remove the problem files one by one. Could you please give us your valuable advice? Thank you. We downloaded about 1 million compounds in pdbq format from VirtualFlow (https://virtual-flow.org/real-library?check_logged_in=1) to use win Ina -GPU for virtual filtering. However, there are some problems with the downloaded database, such as some pdbqt files opened with pymol showing that some atoms are not bonded (see Figure 1 and attachment PV-000093189464_5_T1.pdbqt), Some files contain Si atoms and are therefore identified as illegal pdbqt files (attachment PV-001919877975_1_T1.pdbqt).

Every time you run vina-GPU, if the input ligand file contains one or more problematic files, it will not run correctly, but a large number of compounds cannot be time-consuming and laborious to check and remove the problem files one by one. Could you please give us your valuable advice? Thank you.