
樊樊樊 发表于 2025-8-18 10:14
AF3结果解读:https://alphafoldserver.com/guides#section-3:-interpreting-results-from-alphafold-serve ...
student0618 发表于 2025-9-22 19:04
蛋白-蛋白常用的就是ipTM和pDockQ 分数,不过它们是confidence 相关不是稳定性。结合能应该还是MD算较好。
...
HNUST 发表于 2025-9-22 22:44
亲测,两个不能结合的东西进行af3预测,然后模拟,他不会分离。因为时间太短,平常100ns看不出来
夜航星 发表于 2025-9-23 09:00
请问同志用的GMX进行的模拟吗?预测得到的分数是多少?我尝试过使用AF3预测酶和蛋白质,结果得到的iptm并 ...
HNUST 发表于 2025-9-22 22:44
亲测,两个不能结合的东西进行af3预测,然后模拟,他不会分离。因为时间太短,平常100ns看不出来
夜航星 发表于 2025-9-29 14:18
我是接着我师兄的东西用autodock做某两个物质的分子对接,结果发现不管怎么对接都对接不上,用AF3就能对 ...
夜航星 发表于 2025-9-23 08:42
boltz-2我甚至还加了他们在国外的社群,准备试试呢,可惜老师那边催的急,就用AF3了,而且我们这里的网速 ...
student0618 发表于 2025-9-29 21:59
我另开新帖写了些boltz-2的试用心得,见 http://bbs.keinsci.com/thread-56056-1-1.html
夜航星 发表于 2025-11-7 08:41
感谢同志你的帮助,我目前也正在探索boltz-2的使用,请问能做个视频发到网站上讲讲吗?
student0618 发表于 2025-9-29 21:59
我另开新帖写了些boltz-2的试用心得,见 http://bbs.keinsci.com/thread-56056-1-1.html
夜航星 发表于 2025-11-27 17:18
顺便问一下同志,这里面结合能和iptm哪个更能反映蛋白质之间的结合能力或者说某些有抗菌能力的多肽的MIC ...
Properties (affinity)
properties is an optional field that allows you to specify whether you want to compute the affinity. If enabled, you must also provide the chain_id corresponding to the small molecule against which the affinity will be computed. Only one single small molecule can be specified for affinity computation. It must be a ligand chain (not a protein, DNA or RNA) and has to be at most 128 atoms counting heavy atoms and hydrogens kept by RDKit RemoveHs, however, we do not recommend running the affinity module with ligands significantly larger than 56 atoms (counted as above, limit set during training). At this point, Boltz only supports the computation of affinity of small molecules to protein targets, if ran with an RNA/DNA/co-factor target, the code will not crash but the output will be unreliable.
| 欢迎光临 计算化学公社 (http://bbs.keinsci.com/) | Powered by Discuz! X3.3 |