{"id":1107993,"date":"2024-11-28T20:31:59","date_gmt":"2024-11-29T04:31:59","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-blog-post&p=1107993"},"modified":"2024-11-28T20:32:35","modified_gmt":"2024-11-29T04:32:35","slug":"tamgen","status":"publish","type":"msr-blog-post","link":"https:\/\/www.microsoft.com\/en-us\/research\/articles\/tamgen\/","title":{"rendered":"\u52a0\u901f\u836f\u7269\u53d1\u73b0\uff1a\u57fa\u4e8e\u751f\u6210\u5f0fAI\u7684\u9776\u70b9\u611f\u77e5\u5206\u5b50\u751f\u6210\u5668TamGen"},"content":{"rendered":"\n

\u5168\u7403\u5065\u5eb7\u836f\u7269\u7814\u53d1\u4e2d\u5fc3 \uff08GHDDI\uff09 \u548c \u5fae\u8f6f\u7814\u7a76\u9662\u79d1\u5b66\u667a\u80fd\u4e2d\u5fc3\uff08Microsoft Research AI for Science\uff09\u56e2\u961f\u901a\u8fc7\u5f00\u53d1 TamGen \u5728\u7814\u7a76\u7ed3\u6838\u75c5\u65b9\u9762\u53d6\u5f97\u4e86\u91cd\u8981\u7a81\u7834<\/a>\u3002TamGen \u662f\u4e00\u6b3e\u57fa\u4e8e Transformer \u6a21\u578b\u7684\u5f00\u6e90\u5316\u5b66\u8bed\u8a00\u6a21\u578b\uff0c\u7528\u4e8e\u5f00\u53d1\u7279\u5b9a\u9776\u70b9\u7684\u836f\u7269\u5316\u5408\u7269\u3002\u8054\u5408\u56e2\u961f\u901a\u8fc7\u5bc6\u5207\u5408\u4f5c\u6210\u529f\u786e\u5b9a\u4e86\u51e0\u79cd\u6709\u524d\u666f\u7684\u7ed3\u6838\u75c5\u86cb\u767d\u9176\u6291\u5236\u5242\uff0c\u5176\u4e2d\u6700\u6709\u6548\u7684\u5316\u5408\u7269\u8868\u73b0\u51fa\u663e\u8457\u7684\u751f\u7269\u6d3b\u6027\u3002\u7814\u7a76\u8868\u660e\uff0cTamGen \u8fd8\u53ef\u4ee5\u901a\u8fc7\u8bbe\u8ba1\u9776\u6807\u611f\u77e5\u7247\u6bb5\u6765\u4f18\u5316\u73b0\u6709\u5206\u5b50\uff0c\u4ece\u800c\u53d1\u73b0\u57fa\u4e8e\u5df2\u77e5\u5206\u5b50\u6838\u5fc3\u7ed3\u6784\u7684\u65b0\u578b\u5316\u5408\u7269\u3002\u8be5\u9879\u7814\u7a76\u5df2\u5728\u300a\u81ea\u7136-\u901a\u8baf\u300b\uff08Nature Communications\uff09\u6742\u5fd7\u4e0a\u53d1\u8868\u3002<\/p>\n\n\n\n

\"TamGen<\/figure>\n\n\n\n

TamGen<\/strong><\/p>\n\n\n\n

\u8bba\u6587\u94fe\u63a5\uff1ahttps:\/\/www.nature.com\/articles\/s41467-024-53632-4 (opens in new tab)<\/span><\/a><\/p>\n\n\n\n

\u5f00\u6e90\u94fe\u63a5\uff1ahttps:\/\/github.com\/microsoft\/TamGen (opens in new tab)<\/span><\/a><\/p>\n\n\n\n

\u751f\u6210\u5f0fAI\u52a9\u529b\u7a81\u7834\u836f\u7269\u53d1\u73b0\u7684\u5c40\u9650\u6027<\/h2>\n\n\n\n

\u751f\u6210\u5f0f AI \u901a\u8fc7\u8ba9\u8ba1\u7b97\u673a\u81ea\u4e3b\u5b66\u4e60\u5e76\u751f\u6210\u521b\u65b0\u5185\u5bb9\uff0c\u4e3a\u79d1\u5b66\u63a2\u7d22\u5f00\u8f9f\u4e86\u65b0\u7684\u9053\u8def\u3002TamGen \u5229\u7528\u751f\u6210\u5f0f AI \u521b\u5efa\u65b0\u7684\u5206\u5b50\u5316\u5408\u7269\uff0c\u4e3a\u836f\u7269\u53d1\u73b0\u63d0\u4f9b\u4e86\u4e00\u79cd\u5168\u65b0\u7684\u65b9\u6cd5\u3002\u4e0e\u4f20\u7edf\u4f9d\u8d56\u5b9e\u8bc1\u3001\u7cfb\u7edf\u7b5b\u9009\u5df2\u77e5\u5316\u5408\u7269\u7684\u65b9\u6cd5\u4e0d\u540c\uff0c\u751f\u6210\u5f0f AI \u80fd\u591f\u66f4\u9ad8\u6548\u5730\u63a2\u7d22\u66f4\u5e7f\u6cdb\u7684\u65b0\u578b\u5316\u5408\u7269\uff08\u56fe1\uff09\uff0c\u514b\u670d\u4e86\u4f20\u7edf\u65b9\u6cd5\u5728\u5e9e\u5927\u5316\u5b66\u5e93\u7b5b\u9009\u8fc7\u7a0b\u4e2d\u6240\u9762\u4e34\u7684\u8017\u65f6\u957f\u3001\u590d\u6742\u5ea6\u9ad8\u4e14\u6210\u672c\u6602\u8d35\u7684\u95ee\u9898\u3002<\/p>\n\n\n\n

\"\u4e0e\u4f20\u7edf\u57fa\u4e8e\u7b5b\u9009\u7684\u836f\u7269\u53d1\u73b0\u65b9\u6cd5\u76f8\u6bd4\uff0c\u57fa\u4e8e\u751f\u6210\u5f0f
\u56fe1\uff1a\u4e0e\u4f20\u7edf\u57fa\u4e8e\u7b5b\u9009\u7684\u836f\u7269\u53d1\u73b0\u65b9\u6cd5\u76f8\u6bd4\uff0c\u57fa\u4e8e\u751f\u6210\u5f0f AI \u7684\u65b9\u6cd5\u80fd\u591f\u63a2\u7d22\u65b0\u578b\u5316\u5408\u7269\u3002<\/figcaption><\/figure>\n\n\n\n

TamGen\u5de5\u4f5c\u6d41\u7a0b<\/h2>\n\n\n\n

\u5229\u7528\u751f\u6210\u5f0f AI\uff0cTamGen \u8bbe\u8ba1\u4e86\u7279\u5b9a\u9776\u70b9\u7684\u5316\u5b66\u5316\u5408\u7269\u3002\u57fa\u4e8e\u5927\u8bed\u8a00\u6a21\u578b\uff08LLMs\uff09\u7684\u6210\u529f\uff0c\u7814\u7a76\u5458\u4eec\u91c7\u7528\u4e86\u7c7b\u4f3c\u7684\u6280\u672f\u6765\u751f\u6210\u5316\u5408\u7269\u3002\u9996\u5148\uff0c\u5c06\u5206\u5b50\u8f6c\u6362\u4e3a\u7b80\u5316\u5206\u5b50\u8f93\u5165\u7ebf\u6027\u8868\u8fbe\u7cfb\u7edf\uff08Simplified Molecular Input Line-entry System, SMILES\uff09\uff0c\u8fd9\u662f\u4e00\u79cd\u5c06\u5206\u5b50\u7ed3\u6784\u8868\u793a\u4e3a\u7b26\u53f7\u5e8f\u5217\u7684\u6807\u8bb0\u6cd5\uff0c\u7c7b\u4f3c\u4e8e\u6587\u672c\u3002<\/p>\n\n\n\n

\u4e3a\u4e86\u751f\u6210\u7279\u5b9a\u9776\u70b9\u7684\u5316\u5408\u7269\uff0c\u7814\u7a76\u5458\u4eec\u5f00\u53d1\u4e86\u4e00\u4e2a\u7f16\u7801\u5668\u6765\u5904\u7406\u6709\u5173\u86cb\u767d\u8d28\u7684\u4fe1\u606f\uff0c\u5305\u62ec 3D \u7ed3\u6784\u4fe1\u606f\u3002\u53e6\u5916\uff0c\u4e00\u4e2a\u7ed3\u5408\u4e86\u533b\u5b66\u4e13\u5bb6\u89c1\u89e3\u7684\u4e0a\u4e0b\u6587\u7f16\u7801\u5668\uff0c\u6574\u5408\u4e86\u6709\u5173\u86cb\u767d\u8d28\u9776\u6807\u4ee5\u53ca\u7ecf\u9a8c\u8bc1\u6709\u6548\u6216\u88ab\u786e\u5b9a\u4e3a\u6709\u524d\u666f\u7684\u5316\u5408\u7269\u4fe1\u606f\u3002\u901a\u8fc7\u96c6\u6210\u4e13\u4e1a\u4fe1\u606f\u548c\u8ba1\u7b97\u5206\u6790\uff0c\u8be5\u7f16\u7801\u5668\u53ef\u6307\u5bfc\u5316\u5408\u7269\u751f\u6210\u5668\u751f\u6210\u66f4\u6709\u53ef\u80fd\u4e0e\u86cb\u767d\u8d28\u7ed3\u5408\u7684\u65b0\u5206\u5b50\u3002\u5de5\u4f5c\u6d41\u7a0b\u5982\u56fe2\u6240\u793a\u3002<\/p>\n\n\n\n

\"TamGen
\u56fe2\uff1aTamGen \u7684\u5de5\u4f5c\u6d41\u7a0b<\/figcaption><\/figure>\n\n\n\n

\u901a\u8fc7\u8ba1\u7b97\u6027\u80fd\u8bc4\u4f30TamGen<\/h2>\n\n\n\n

\u4e3a\u4e86\u8bc4\u4f30 TamGen \u7684\u6027\u80fd\uff0c\u7814\u7a76\u5458\u4eec\u5c06\u5176\u4e0e\u5176\u4ed6\u4e94\u79cd\u5e38\u89c1\u7684\u65b9\u6cd5\u8fdb\u884c\u4e86\u6bd4\u8f83\uff0c\u8fd9\u4e9b\u65b9\u6cd5\u5747\u7528\u4e8e\u521b\u5efa\u4e0e\u7279\u5b9a\u86cb\u767d\u8d28\u7ed3\u5408\u7684 3D \u7ed3\u6784\u5206\u5b50\u3002\u5229\u7528 CrossDocked \u57fa\u51c6\u6570\u636e\u96c6\uff0c\u7814\u7a76\u5458\u4eec\u8bc4\u4f30\u4e86\u8fd9\u4e9b\u65b9\u6cd5\uff0c\u8be5\u6570\u636e\u96c6\u7528\u4e8e\u8bc4\u4f30\u5728 AI \u7814\u7a76\u4e2d\u4ee5\u9776\u86cb\u767d\u4e3a\u6761\u4ef6\u7684\u5206\u5b50\u751f\u6210\u8d28\u91cf\u3002<\/p>\n\n\n\n

\u8bc4\u4f30\u6307\u6807\u5305\u62ec\uff1a<\/p>\n\n\n\n

    \n
  • Docking score<\/strong>\uff1a\u8bc4\u4f30\u5206\u5b50\u4e0e\u76ee\u6807\u86cb\u767d\u7684\u7ed3\u5408\u7a0b\u5ea6\u3002<\/li>\n\n\n\n
  • \u836f\u7269\u76f8\u4f3c\u6027\u5b9a\u91cf\u4f30\u8ba1 \uff08Quantitative Estimate of Drug-likeness, QED\uff09<\/strong>\uff1a\u8bc4\u4f30\u5206\u5b50\u7684\u5019\u9009\u836f\u7269\u8d28\u91cf\u3002<\/li>\n\n\n\n
  • \u5408\u6210\u53ef\u53ca\u6027\u5206\u6570 \uff08Synthesis Accessibility Score, SAS\uff09<\/strong>\uff1a\u6d4b\u91cf\u5728\u5b9e\u9a8c\u5ba4\u4e2d\u5408\u6210\u7279\u5b9a\u5316\u5408\u7269\u7684\u96be\u6613\u7a0b\u5ea6\u3002<\/li>\n\n\n\n
  • \u7c7b\u836f\u4e94\u539f\u5219\uff08Lipinski’s rule of five,\u00a0Ro5\uff09<\/strong>\uff1a\u5224\u5b9a\u5316\u5408\u7269\u5f00\u53d1\u6210\u53e3\u670d\u836f\u7269\u7684\u53ef\u80fd\u6027\u3002<\/li>\n\n\n\n
  • LogP<\/strong>\uff1a\u6d4b\u8bd5\u5316\u5408\u7269\u5728\u6c34\u548c\u8102\u80aa\u4e4b\u95f4\u5206\u914d\u7684\u80fd\u529b\u3002<\/li>\n\n\n\n
  • \u591a\u6837\u6027<\/strong>\uff1a\u6d4b\u91cf\u5316\u5408\u7269\u96c6\u5408\u4e2d\u4e0d\u540c\u5206\u5b50\u7ed3\u6784\u548c\u6027\u8d28\u7684\u8303\u56f4\u3002<\/li>\n<\/ul>\n\n\n\n

    \u56fe3\u5c55\u793a\u4e86 TamGen \u6574\u4f53\u6027\u80fd\u7684\u8868\u73b0\u3002\u867d\u7136\u5176\u4ed6\u65b9\u6cd5\u4e5f\u53ef\u4ee5\u4ea7\u751f\u5177\u6709\u5f3a\u7ed3\u5408\u80fd\u529b\u7684\u5316\u5408\u7269\uff0c\u4f46\u5b83\u4eec\u901a\u5e38\u5305\u62ec\u591a\u4e2a\u4e92\u8fde\u7684\u73af\u72b6\u7ed3\u6784\u3002\u7814\u7a76\u8868\u660e\uff0c\u8fd9\u4e9b\u7ed3\u6784\u7684\u589e\u52a0\u4f1a\u964d\u4f4e\u5408\u6210\u53ef\u53ca\u6027\uff08SAS\uff09\u5e76\u589e\u52a0\u7ec6\u80de\u6bd2\u6027\uff0c\u4f7f\u8fd9\u4e9b\u5316\u5408\u7269\u66f4\u96be\u5f00\u53d1\u3002\u7814\u7a76\u5458\u4eec\u8ba4\u4e3a\uff0c\u9884\u8bad\u7ec3\u7b56\u7565\u6709\u52a9\u4e8e\u63d0\u9ad8 TamGen \u751f\u6210\u7684\u5316\u5408\u7269\u7684\u6574\u4f53\u6709\u6548\u6027\u3002<\/p>\n\n\n\n

    \"TamGen
    \u56fe3\uff1aTamGen \u8ba1\u7b97\u6027\u80fd\u9a8c\u8bc1\u7ed3\u679c<\/figcaption><\/figure>\n\n\n\n

    TamGen\u5b9e\u9a8c\u9a8c\u8bc1<\/h2>\n\n\n\n

    \u4e3a\u4e86\u786e\u4fdd\u7814\u7a76\u7684\u5b9e\u9645\u5e94\u7528\uff0c\u7814\u7a76\u5458\u4eec\u8fd8\u5728\u6e7f\u5b9e\u9a8c\u73af\u5883\u4e2d\u8fdb\u884c\u4e86\u9a8c\u8bc1\u3002\u5728\u6e7f\u5b9e\u9a8c\u4e2d\uff0c\u7814\u7a76\u5458\u4eec\u9009\u62e9\u4e86\u7ed3\u6838\u5206\u679d\u6746\u83cc\u4e2d\u7684 ClpP \u86cb\u767d\u9176\u4f5c\u4e3a\u9776\u6807\uff0c\u8be5\u86cb\u767d\u9176\u5bf9\u7ec6\u83cc\u5728\u538b\u529b\u6761\u4ef6\u4e0b\u7684\u5b58\u6d3b\u8d77\u7740\u91cd\u8981\u4f5c\u7528\u3002\u901a\u8fc7\u8bbe\u8ba1\u3001\u4f18\u5316\u548c\u6d4b\u8bd5\u4e09\u4e2a\u9636\u6bb5\uff0cTamGen \u53ef\u6709\u6548\u8bc6\u522b\u7528\u4e8e\u7ed3\u6838\u75c5\u836f\u7269\u53d1\u73b0\u7684\u5206\u5b50\u5316\u5408\u7269\uff08\u5982\u56fe4\u6240\u793a\uff09\u3002<\/p>\n\n\n\n

    \u8bbe\u8ba1\u9636\u6bb5<\/strong>\uff1a\u9996\u5148\u4f7f\u7528 TamGen \u5206\u6790\u86cb\u767d\u9176\u7684\u7ed3\u5408\u53e3\u888b\uff0c\u5176\u4ed6\u5206\u5b50\u53ef\u4ee5\u7531\u6b64\u9644\u7740\u5728\u86cb\u767d\u9176\u4e0a\u5e76\u5f71\u54cd\u5176\u529f\u80fd\u3002TamGen \u751f\u6210\u4e86\u5927\u7ea62,600\u79cd\u53ef\u4ee5\u653e\u5165\u8fd9\u4e2a\u53e3\u888b\u7684\u6f5c\u5728\u5316\u5408\u7269\uff0c\u6839\u636e\u5b83\u4eec\u4e0e\u86cb\u767d\u9176\u7684\u7ed3\u5408\u7a0b\u5ea6\u53ca\u5176\u9884\u6d4b\u7684\u751f\u7269\u5b66\u6548\u5e94\uff0c\u7814\u7a76\u5458\u4eec\u5bf9\u8fd9\u4e9b\u5316\u5408\u7269\u8fdb\u884c\u4e86\u8bc4\u4f30\uff0c\u5e76\u5c06\u8303\u56f4\u7f29\u5c0f\u5230\u4e864\u4e2a\u76ee\u6807\u5019\u9009\u5316\u5408\u7269\u3002<\/p>\n\n\n\n

    \u4f18\u5316\u9636\u6bb5<\/strong>\uff1a\u63a5\u4e0b\u6765\uff0c\u7814\u7a76\u5458\u4eec\u5c06\u8bbe\u8ba1\u9636\u6bb5\u9274\u5b9a\u76844\u79cd\u5316\u5408\u7269\u4ee5\u53ca\u4e4b\u524d\u5728\u5b9e\u9a8c\u5ba4\u5b9e\u9a8c\u4e2d\u9a8c\u8bc1\u76843\u4e2a\u5206\u5b50\u7247\u6bb5\u8f93\u5165\u5230 TamGen \u4e2d\uff0c\u8fd9\u4e00\u8fc7\u7a0b\u603b\u5171\u751f\u6210\u4e868,600\u79cd\u65b0\u5316\u5408\u7269\u3002\u4f7f\u7528\u76f8\u540c\u7684\u6807\u51c6\uff0c\u7814\u7a76\u5458\u4eec\u518d\u6b21\u7b5b\u9009\u4e86\u8fd9\u4e9b\u5316\u5408\u7269\uff0c\u6700\u7ec8\u5c06\u9009\u62e9\u8303\u56f4\u7f29\u5c0f\u5230\u4e86296\u79cd\u5316\u5408\u7269\u3002<\/p>\n\n\n\n

    \u6d4b\u8bd5\u9636\u6bb5<\/strong>\uff1a\u7531\u4e8e\u5408\u6210\u6240\u6709296\u79cd\u5316\u5408\u7269\u5e76\u4e0d\u73b0\u5b9e\uff0c\u6240\u4ee5\u7814\u7a76\u5458\u4eec\u4ece\u5546\u4e1a\u5316\u5b66\u5e93\u4e2d\u7b5b\u9009\u4e86\u7c7b\u4f3c\u7684\u5316\u5408\u7269\uff0c\u5e76\u6d4b\u8bd5\u4e86\u5b83\u4eec\u5bf9\u7ed3\u6838\u75c5\u7684\u521d\u59cb\u6d3b\u6027\u3002\u5176\u4e2d5\u79cd\u5316\u5408\u7269\u663e\u793a\u51fa\u4e86\u6709\u524d\u666f\u7684\u7ed3\u679c\u3002\u968f\u540e\uff0c\u7814\u7a76\u5458\u4eec\u5408\u6210\u4e86\u5176\u4e2d\u4e00\u79cd\u539f\u59cb\u5316\u5408\u7269\u4ee5\u53ca\u53e6\u4e00\u79cd\u5316\u5408\u7269\u7684\u4e24\u4e2a\u53d8\u4f53\u3002\u6b64\u5916\uff0c\u7814\u7a76\u5458\u4eec\u8fd8\u5c06\u751f\u6210\u7684\u5316\u5408\u7269\u5206\u4e3a\u4e86\u82e5\u5e72\u7c07\uff0c\u5e76\u6839\u636e\u5bf9\u63a5\u8bc4\u5206\u4ece\u6bcf\u7c07\u4e2d\u9009\u51fa\u524d10%\uff0c\u7ecf\u8fc7\u4eba\u5de5\u5ba1\u6838\uff0c\u53c8\u5408\u6210\u4e868\u79cd\u5316\u5408\u7269\u3002<\/p>\n\n\n\n

    \u5fae\u8f6f\u7814\u7a76\u9662\u7684\u56e2\u961f\u4f7f\u7528 TamGen \u751f\u6210\u4e86\u5019\u9009\u5316\u5408\u7269\u4e4b\u540e\uff0cGHDDI \u56e2\u961f\u5bf9\u6b64\u8fdb\u884c\u4e86\u7ed3\u5408\u5206\u6790\u3001\u7ed3\u6784-\u6d3b\u6027\u5173\u7cfb\u7814\u7a76\u4ee5\u53ca\u5b9e\u9a8c\u5ba4\u5b9e\u9a8c\uff0c\u4ee5\u9a8c\u8bc1\u8fd9\u4e9b\u5316\u5408\u7269\u5bf9 ClpP \u86cb\u767d\u9176\u7684\u6291\u5236\u6548\u679c\uff0c\u5e76\u786e\u5b9a\u4e86\u5176 IC50 \u503c\u2014\u2014\u8fd9\u4e00\u6307\u6807\u8861\u91cf\u4e86\u6291\u523650%\u9176\u6d3b\u6027\u6240\u9700\u7684\u5316\u5408\u7269\u6d53\u5ea6\uff0cIC50 \u503c\u8d8a\u4f4e\uff0c\u6548\u80fd\u8d8a\u5f3a\u3002\u5728\u6d4b\u8bd5\u768416\u79cd\u5316\u5408\u7269\u4e2d\uff0c\u670914\u79cd\u8868\u73b0\u51fa\u5f88\u5f3a\u7684\u6291\u5236\u6d3b\u6027\uff0cIC50 \u503c\u5747\u4f4e\u4e8e40 \u03bcM\uff0c\u663e\u793a\u51fa\u826f\u597d\u7684\u6f5c\u529b\u3002\u5176\u4e2d\u6700\u6709\u6548\u7684\u5316\u5408\u7269 IC50 \u503c\u4e3a1.88 \u03bcM\u3002<\/p>\n\n\n\n

    \"TamGen
    \u56fe4\uff1aTamGen \u6e7f\u5b9e\u9a8c\u5ba4\u9a8c\u8bc1\u8fc7\u7a0b<\/figcaption><\/figure>\n\n\n\n

    \u4ece\u5206\u5b50\u5230\u7247\u6bb5\u751f\u6210<\/h2>\n\n\n\n

    \u9664\u4e86\u751f\u6210\u5168\u65b0\u7684\u5206\u5b50\uff0cTamGen \u8fd8\u80fd\u591f\u901a\u8fc7\u8bbe\u8ba1\u5206\u5b50\u5316\u5408\u7269\u7684\u8f83\u5c0f\u90e8\u5206\u6765\u4f18\u5316\u73b0\u6709\u5316\u5408\u7269\u3002\u5728\u7247\u6bb5\u751f\u6210\u8fc7\u7a0b\u4e2d\uff0c\u82e5\u7ed9\u5b9a\u4e00\u4e2a\u86cb\u767d\u8d28\u9776\u6807\u548c\u4e00\u4e2a\u9700\u4fdd\u7559\u7684\u7279\u5b9a\u5206\u5b50\u7ed3\u6784\uff0cTamGen \u5c31\u53ef\u4ee5\u56f4\u7ed5\u8be5\u7ed3\u6784\u8bbe\u8ba1\u5168\u65b0\u7684\u5316\u5408\u7269\u3002TamGen \u5e76\u975e\u4ece\u73b0\u6709\u7684\u5316\u5408\u7269\u5e93\u4e2d\u9009\u62e9\u7247\u6bb5\uff0c\u800c\u662f\u8bbe\u8ba1\u4e86\u53ef\u8bc6\u522b\u9776\u6807\u7684\u7247\u6bb5\uff0c\u8fd9\u610f\u5473\u7740\u5b83\u4eec\u5305\u542b\u4e0e\u9776\u86cb\u767d\u76f8\u5173\u7684\u4fe1\u606f\u3002\u8fd9\u79cd\u65b9\u6cd5\u8d85\u8d8a\u4e86\u4f20\u7edf\u4f9d\u8d56\u4e8e\u9884\u5148\u5b58\u5728\u7684\u6570\u636e\u5e93\u7684\u65b9\u5f0f\uff0c\u6253\u7834\u4e86\u4f20\u7edf\u65b9\u5f0f\u5bf9\u65b0\u9896\u6027\u548c\u6709\u6548\u6027\u7684\u9650\u5236\u3002<\/p>\n\n\n\n

    \u6b64\u5916\uff0c\u7814\u7a76\u5458\u4eec\u8fd8\u8c03\u6574\u4e86 TamGen \u5316\u5408\u7269\u751f\u6210\u5668\u7684\u8f93\u5165\uff0c\u8bbe\u8ba1\u4e86\u76ee\u6807\u611f\u77e5\u7247\u6bb5\u3002\u901a\u8fc7\u91cd\u5199 SMILES \u5b57\u7b26\u4e32\uff0c\u6765\u786e\u5b9a\u4fdd\u7559\u7684\u7247\u6bb5\u548c\u8fdb\u4e00\u6b65\u589e\u957f\u7684\u7279\u5b9a\u8fde\u63a5\u70b9\uff0c\u4ece\u800c\u786e\u4fdd\u5176\u5728\u6240\u9700\u7684\u589e\u957f\u7ad9\u70b9\u7ec8\u6b62\u3002\u7136\u540e\uff0c\u518d\u5c06\u4fee\u6539\u540e\u7684 SMILES \u5b57\u7b26\u4e32\u8f93\u5165TamGen \u7684\u5316\u5408\u7269\u751f\u6210\u5668\u8fdb\u884c\u6269\u5c55\u3002<\/p>\n\n\n\n

    \u4e3a\u4e86\u8bc4\u4f30\u8fd9\u79cd\u65b9\u6cd5\uff0c\u5fae\u8f6f\u7814\u7a76\u9662\u4e0e GHDDI \u5408\u4f5c\uff0c\u9488\u5bf9\u7ed3\u6838\u75c5 \u7684 ClpP \u86cb\u767d\u9176\u8fdb\u884c\u4e86\u6e7f\u5b9e\u9a8c\uff0c\u53d1\u73b0\u4e0e\u539f\u59cb\u5316\u5408\u7269\u76f8\u6bd4\uff0cTamGen \u751f\u6210\u7684\u5206\u5b50\u5316\u5408\u7269\u7684\u7ed3\u5408\u4eb2\u548c\u529b\u63d0\u9ad8\u4e86\u5341\u500d\u4ee5\u4e0a\u3002\u800c\u4e14\uff0cTamGen \u8bbe\u8ba1\u7684\u90e8\u5206\u5316\u5408\u7269\u8fd8\u8868\u73b0\u51fa\u6162\u7ed3\u5408\uff08slow binding\uff09\uff0c\u8868\u660e\u5b83\u4eec\u5177\u6709\u6301\u4e45\u4f5c\u7528\u7684\u6f5c\u529b\uff0c\u5e76\u53ef\u4ee5\u63d0\u9ad8\u5bf9\u76ee\u6807\u9776\u86cb\u767d\u7684\u9009\u62e9\u6027\u3002<\/p>\n\n\n\n

    \u4eba\u5de5\u667a\u80fd\u5728\u836f\u7269\u53d1\u73b0\u4e2d\u7684\u6f5c\u80fd<\/h2>\n\n\n\n

    \u901a\u8fc7\u5c06\u5148\u8fdb\u7684\u5206\u5b50\u5efa\u6a21\u7cfb\u7edf\u4e0e\u79d1\u7814\u4eba\u5458\u548c AI \u7cfb\u7edf\u7684\u534f\u4f5c\u76f8\u7ed3\u5408\uff0cTamGen \u5145\u5206\u5c55\u793a\u4e86\u751f\u6210\u5f0f AI \u5728\u73b0\u5b9e\u836f\u7269\u8bbe\u8ba1\u4e2d\u7684\u5de8\u5927\u6f5c\u529b\u3002\u5728\u4f20\u7edf\u65b9\u6cd5\u4e2d\u9700\u8981\u6570\u5e74\u624d\u80fd\u5b8c\u6210\u7684\u5de5\u4f5c\uff0c\u73b0\u5728\u53ef\u5728\u5f88\u77ed\u7684\u65f6\u95f4\u5185\u5b9e\u73b0\u3002\u8fd9\u9879\u7814\u7a76\u7a81\u663e\u4e86\u4eba\u5de5\u667a\u80fd\u5728\u836f\u7269\u53d1\u73b0\u4e2d\u65e5\u76ca\u91cd\u8981\u7684\u4f5c\u7528\uff0c\u5e76\u6709\u671b\u4e3a\u7ed3\u6838\u75c5\u7b49\u6301\u7eed\u6027\u4f20\u67d3\u75c5\u63d0\u4f9b\u6709\u6548\u7684\u89e3\u51b3\u65b9\u6848\u3002<\/p>\n\n\n\n

    \u5c55\u671b\u672a\u6765\uff0c\u5fae\u8f6f\u7814\u7a76\u9662\u8ba1\u5212\u5c06\u66f4\u5148\u8fdb\u7684\u6280\u672f\u96c6\u6210\u5230 TamGen \u4e2d\uff0c\u5305\u62ec\u7528\u4e8e\u751f\u6210 3D \u7ed3\u6784\u7684\u6269\u6563\u6a21\u578b\u3001\u5e94\u7528\u7269\u7406\u7ea6\u675f\u7684\u5f3a\u5316\u5b66\u4e60\uff0c\u4ee5\u53ca\u6355\u83b7\u86cb\u767d\u8d28\u5f62\u6001\u53d8\u5316\u7684\u5206\u5b50\u52a8\u529b\u5b66\u6a21\u62df\u3002\u8fd9\u4e9b\u589e\u5f3a\u529f\u80fd\u5c06\u6709\u52a9\u4e8e\u6539\u5584\u751f\u6210\u7684\u5206\u5b50\u5316\u5408\u7269\u4e0e\u9776\u86cb\u767d\u7684\u7ed3\u5408\u7a0b\u5ea6\uff0c\u63d0\u9ad8\u5408\u6210\u53ef\u884c\u6027\u5e76\u589e\u5f3a\u5176\u4ed6\u5173\u952e\u836f\u7269\u7279\u6027\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"

    \u5168\u7403\u5065\u5eb7\u836f\u7269\u7814\u53d1\u4e2d\u5fc3 \uff08GHDDI\uff09 \u548c \u5fae\u8f6f\u7814\u7a76\u9662\u79d1\u5b66\u667a\u80fd\u4e2d\u5fc3\uff08Microsoft Research AI for Science\uff09\u56e2\u961f\u901a\u8fc7\u5f00\u53d1 TamGen \u5728\u7814\u7a76\u7ed3\u6838\u75c5\u65b9\u9762\u53d6\u5f97\u4e86\u91cd\u8981\u7a81\u7834\u3002TamGen \u662f\u4e00\u6b3e\u57fa\u4e8e Transformer \u6a21\u578b\u7684\u5f00\u6e90\u5316\u5b66\u8bed\u8a00\u6a21\u578b\uff0c\u7528\u4e8e\u5f00\u53d1\u7279\u5b9a\u9776\u70b9\u7684\u836f\u7269\u5316\u5408\u7269\u3002\u8054\u5408\u56e2\u961f\u901a\u8fc7\u5bc6\u5207\u5408\u4f5c\u6210\u529f\u786e\u5b9a\u4e86\u51e0\u79cd\u6709\u524d\u666f\u7684\u7ed3\u6838\u75c5\u86cb\u767d\u9176\u6291\u5236\u5242\uff0c\u5176\u4e2d\u6700\u6709\u6548\u7684\u5316\u5408\u7269\u8868\u73b0\u51fa\u663e\u8457\u7684\u751f\u7269\u6d3b\u6027\u3002\u7814\u7a76\u8868\u660e\uff0cTamGen \u8fd8\u53ef\u4ee5\u901a\u8fc7\u8bbe\u8ba1\u9776\u6807\u611f\u77e5\u7247\u6bb5\u6765\u4f18\u5316\u73b0\u6709\u5206\u5b50\uff0c\u4ece\u800c\u53d1\u73b0\u57fa\u4e8e\u5df2\u77e5\u5206\u5b50\u6838\u5fc3\u7ed3\u6784\u7684\u65b0\u578b\u5316\u5408\u7269\u3002\u8be5\u9879\u7814\u7a76\u5df2\u5728\u300a\u81ea\u7136-\u901a\u8baf\u300b\uff08Nature Communications\uff09\u6742\u5fd7\u4e0a\u53d1\u8868\u3002 TamGen \u8bba\u6587\u94fe\u63a5\uff1ahttps:\/\/www.nature.com\/articles\/s41467-024-53632-4 (opens in new tab) \u5f00\u6e90\u94fe\u63a5\uff1ahttps:\/\/github.com\/microsoft\/TamGen (opens in new tab) \u751f\u6210\u5f0f AI \u901a\u8fc7\u8ba9\u8ba1\u7b97\u673a\u81ea\u4e3b\u5b66\u4e60\u5e76\u751f\u6210\u521b\u65b0\u5185\u5bb9\uff0c\u4e3a\u79d1\u5b66\u63a2\u7d22\u5f00\u8f9f\u4e86\u65b0\u7684\u9053\u8def\u3002TamGen \u5229\u7528\u751f\u6210\u5f0f AI \u521b\u5efa\u65b0\u7684\u5206\u5b50\u5316\u5408\u7269\uff0c\u4e3a\u836f\u7269\u53d1\u73b0\u63d0\u4f9b\u4e86\u4e00\u79cd\u5168\u65b0\u7684\u65b9\u6cd5\u3002\u4e0e\u4f20\u7edf\u4f9d\u8d56\u5b9e\u8bc1\u3001\u7cfb\u7edf\u7b5b\u9009\u5df2\u77e5\u5316\u5408\u7269\u7684\u65b9\u6cd5\u4e0d\u540c\uff0c\u751f\u6210\u5f0f AI \u80fd\u591f\u66f4\u9ad8\u6548\u5730\u63a2\u7d22\u66f4\u5e7f\u6cdb\u7684\u65b0\u578b\u5316\u5408\u7269\uff08\u56fe1\uff09\uff0c\u514b\u670d\u4e86\u4f20\u7edf\u65b9\u6cd5\u5728\u5e9e\u5927\u5316\u5b66\u5e93\u7b5b\u9009\u8fc7\u7a0b\u4e2d\u6240\u9762\u4e34\u7684\u8017\u65f6\u957f\u3001\u590d\u6742\u5ea6\u9ad8\u4e14\u6210\u672c\u6602\u8d35\u7684\u95ee\u9898\u3002 \u5229\u7528\u751f\u6210\u5f0f AI\uff0cTamGen \u8bbe\u8ba1\u4e86\u7279\u5b9a\u9776\u70b9\u7684\u5316\u5b66\u5316\u5408\u7269\u3002\u57fa\u4e8e\u5927\u8bed\u8a00\u6a21\u578b\uff08LLMs\uff09\u7684\u6210\u529f\uff0c\u7814\u7a76\u5458\u4eec\u91c7\u7528\u4e86\u7c7b\u4f3c\u7684\u6280\u672f\u6765\u751f\u6210\u5316\u5408\u7269\u3002\u9996\u5148\uff0c\u5c06\u5206\u5b50\u8f6c\u6362\u4e3a\u7b80\u5316\u5206\u5b50\u8f93\u5165\u7ebf\u6027\u8868\u8fbe\u7cfb\u7edf\uff08Simplified Molecular Input Line-entry System, SMILES\uff09\uff0c\u8fd9\u662f\u4e00\u79cd\u5c06\u5206\u5b50\u7ed3\u6784\u8868\u793a\u4e3a\u7b26\u53f7\u5e8f\u5217\u7684\u6807\u8bb0\u6cd5\uff0c\u7c7b\u4f3c\u4e8e\u6587\u672c\u3002 \u4e3a\u4e86\u751f\u6210\u7279\u5b9a\u9776\u70b9\u7684\u5316\u5408\u7269\uff0c\u7814\u7a76\u5458\u4eec\u5f00\u53d1\u4e86\u4e00\u4e2a\u7f16\u7801\u5668\u6765\u5904\u7406\u6709\u5173\u86cb\u767d\u8d28\u7684\u4fe1\u606f\uff0c\u5305\u62ec 3D \u7ed3\u6784\u4fe1\u606f\u3002\u53e6\u5916\uff0c\u4e00\u4e2a\u7ed3\u5408\u4e86\u533b\u5b66\u4e13\u5bb6\u89c1\u89e3\u7684\u4e0a\u4e0b\u6587\u7f16\u7801\u5668\uff0c\u6574\u5408\u4e86\u6709\u5173\u86cb\u767d\u8d28\u9776\u6807\u4ee5\u53ca\u7ecf\u9a8c\u8bc1\u6709\u6548\u6216\u88ab\u786e\u5b9a\u4e3a\u6709\u524d\u666f\u7684\u5316\u5408\u7269\u4fe1\u606f\u3002\u901a\u8fc7\u96c6\u6210\u4e13\u4e1a\u4fe1\u606f\u548c\u8ba1\u7b97\u5206\u6790\uff0c\u8be5\u7f16\u7801\u5668\u53ef\u6307\u5bfc\u5316\u5408\u7269\u751f\u6210\u5668\u751f\u6210\u66f4\u6709\u53ef\u80fd\u4e0e\u86cb\u767d\u8d28\u7ed3\u5408\u7684\u65b0\u5206\u5b50\u3002\u5de5\u4f5c\u6d41\u7a0b\u5982\u56fe2\u6240\u793a\u3002 \u4e3a\u4e86\u8bc4\u4f30 TamGen \u7684\u6027\u80fd\uff0c\u7814\u7a76\u5458\u4eec\u5c06\u5176\u4e0e\u5176\u4ed6\u4e94\u79cd\u5e38\u89c1\u7684\u65b9\u6cd5\u8fdb\u884c\u4e86\u6bd4\u8f83\uff0c\u8fd9\u4e9b\u65b9\u6cd5\u5747\u7528\u4e8e\u521b\u5efa\u4e0e\u7279\u5b9a\u86cb\u767d\u8d28\u7ed3\u5408\u7684 3D \u7ed3\u6784\u5206\u5b50\u3002\u5229\u7528 CrossDocked \u57fa\u51c6\u6570\u636e\u96c6\uff0c\u7814\u7a76\u5458\u4eec\u8bc4\u4f30\u4e86\u8fd9\u4e9b\u65b9\u6cd5\uff0c\u8be5\u6570\u636e\u96c6\u7528\u4e8e\u8bc4\u4f30\u5728 AI \u7814\u7a76\u4e2d\u4ee5\u9776\u86cb\u767d\u4e3a\u6761\u4ef6\u7684\u5206\u5b50\u751f\u6210\u8d28\u91cf\u3002 \u8bc4\u4f30\u6307\u6807\u5305\u62ec\uff1a […]<\/p>\n","protected":false},"author":34512,"featured_media":1105698,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-content-parent":1012650,"msr_hide_image_in_river":null,"footnotes":""},"research-area":[13556],"msr-locale":[268881],"msr-post-option":[],"class_list":["post-1107993","msr-blog-post","type-msr-blog-post","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","msr-locale-zh_cn"],"msr_assoc_parent":{"id":1012650,"type":"lab"},"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/1107993","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-blog-post"}],"author":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/users\/34512"}],"version-history":[{"count":4,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/1107993\/revisions"}],"predecessor-version":[{"id":1108020,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/1107993\/revisions\/1108020"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/1105698"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1107993"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1107993"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1107993"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=1107993"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}