ÎÄÕÂÎÊÌ⣺The fusion of multi-omics profile and multimodal EEG data contributes to the personalized diagnostic strategy for neurocognitive disorders
½ÒÏþʱ¼ä£º2024Äê1ÔÂ29ÈÕ ½ÒÏþÆÚ¿¯£ºMicrobiome Ó°ÏìÒò×Ó£º16.8 Éñ¾ÈÏÖªÕÏ°(Neurocognitive Disorders, NCDs)ÊÇÒ»ÖÖÓ°ÏìÒ»Ñùƽ³£ÉúÑÄÖÊÁ¿µÄ·ÇѬȾÐÔ¼²²¡£¬È«Çò»¼ÕßÔ¼ÓÐ5000ÍòÈË¡£¶øÏÖÔÚµÄͨÀýÕï¶ÏÒªÁìȱ·¦Ã÷È·µÄÉúÎï±ê¼ÇÎï»òÁîÈËÐÅ·þµÄ¼ì²é¡£Òò´Ë¿ª·¢Õë¶Ô·ÇѬȾÐÔ¼²²¡µÄ¸öÐÔ»¯Õï¶ÏÕ½ÂÔ³ÉΪµ±ÎñÖ®¼±¡£ÏÖÔÚ¶à×éѧÑо¿ºÍÂÑ°×ÖÊ×éѧµÄÆÊÎöÅú×¢£¬¶à×éѧÆÊÎöÔÚÑо¿·ÇѬȾÐÔ¼²²¡ºÍʶ±ðDZÔÚÉúÎï±ê¼ÇÎï·½Ãæ¾ßÓÐDZÁ¦¡£±ðµÄ£¬¶àģ̬ÄÔµçͼ(electroencephalography, EEG)¿ÉÒÔÌṩÓйشóÄÔÔ˶¯µÄʱ¼äÎÂ˳ÐòµÄÐÅÏ¢£¬¿ÉÒÔ¸üÖÜÈ«µØÏàʶ´óÄÔ¹¦Ð§ºÍÅþÁ¬ÐÔ£¬½üÄêÀ´Ô½À´Ô½¶àµØÓÃÓÚÈÏÖªÕÏ°µÄÕï¶Ï¡£»ùÓÚÕâЩÑо¿»ù´¡£¬Ñо¿Ö°Ô±Í¨¹ýÑÏ¿áµÄÆ¥ÅäÐÐÁн«»¼ÓзÇѬȾÐÔ¼²²¡µÄÍíÄêÈËÓëÕý³£ÐàÂõµÄÍíÄêÈ˾ÙÐÐÁ˽ÏÁ¿£¬²¢ÔÚСÊóÖоÙÐÐÁË·à±ã΢ÉúÎïȺÒÆÖ²(Faecal microbiota transplantation, FMT)ÑéÖ¤¡£×îºóÓ¦ÓûúеѧϰËã·¨ÕûºÏ¶à×éѧºÍ¶àģ̬ÄÔµçͼÊý¾Ý£¬¿ª·¢ÐµÄÕ¹ÍûÄ£×Ó£¬ÒÔ׼ȷÇø·Ö·ÇѬȾÐÔ¼²²¡»¼ÕߺͼÓÈë·ÇѬȾÐÔ¼²²¡¸öÐÔ»¯Õï¶ÏµÄ¼ÓÈëÕß¡£
±¾Ñо¿¹²Ñ¡È¡ÁË400λÍíÄêÈ˾ÙÐÐÊÔÑ飬ÆäÖÐÓÐ57Ãû¼ÓÈëÕßÒòÐÅÏ¢²»ÍêÕû¶ø±»É¨³ý¡£Æ¾Ö¤Ïã¸Û°æ±¾µÄÃÉÌØÀû¶ûÈÏÖªÆÀ¹À(Hong Kong version of Montreal Cognitive Assessment, MoCA-HK)µÄ¸÷¸öÁìÓòµÄÈÏÖª¹¦Ð§£¬ÍíÄêÈ˱»·ÖΪÁ½×é¡£ÈôÊǸöÌåµÄMoCA-HKµÃ·ÖµÍÓÚÓëÆäÄêËêºÍ½ÌÓýˮƽÏàÆ¥ÅäµÄͬ°éµÄ16%£¬ÄÇô¸ÃÍíÄêÈ˽«±»»®Èë¡°·ÇѬȾÐÔ¼²²¡×顱(neurocognitive disorders, NCDs)£¬²»È»½«±»»®Èë¡°Õý³£ÐàÂõ×顱(normal aging, NA)¡£
»®·ÖÈ¡·à±ã¾ÙÐкê»ùÒò×éºÍ·Ç°ÐÏò´úл×éѧÆÊÎö£¬ÄòÒºÍâÃÚÌå¾ÙÐÐÂÑ°×ÖÊ×éѧÆÊÎö¡£Ñ¡Ôñ12¸öÔ´óµÄ´ÆÐÔСÊó×÷Ϊģ×Ó£¬ÔÚÌض¨µÄÎÞ²¡ÔÌåÌõ¼þÏÂËÇÑø£¬·à±ã΢ÉúÎïȺÒÆÖ²Ö®ºó£¬¾ÙÐÐÐÐΪʵÑ飬Ëæºó¾ÙÐкê»ùÒò×éºÍ´úл×éѧÆÊÎö¡£Ê¹Óó¦µÀ΢ÉúÎïȺ¡¢ÄÔµçͼÌØÕ÷¡¢´úл×éѧºÍÁÙ´²Êý¾Ý£¬À´ÑµÁ·»úеѧϰģ×Ó¡£ »ùÓÚÄԵ繦ÂÊÆ×ÃܶÈÆÊÎö·¢Ã÷£¬NCDs×éÖЦÁºÍ¦ÂƵ´øµÄÏà¶Ô¹¦ÂÊÆ×ÃܶÈ(power spectral density, PSD)ÏÔÖø½µµÍ£¨Í¼1£©¡£»ùÓÚÄÔµç΢¹Û״̬ÆÊÎö·¢Ã÷£¬NA×éºÍNCDs×éÔÚ΢¹Û״̬ÊôÐÔÉÏÓÐÏÔÖø²î±ð£¨Í¼2£©¡£±ðµÄ£¬½«ÄêËê×÷Ϊ¹²±äÁ¿¾ÙÐпØÖƵÄÇ°ºó£¬·¢Ã÷΢״̬ת»»µÄ¸ÅÂʽϸߡ£ ͼ1 NA×éºÍNCDs×éÖеÄEEG PSDºÍEEG΢״̬ÆÊÎö ͼ2 NA×éºÍNCDs×éµÄ΢״̬A-DÊôÐÔͳ¼Æ ºê»ùÒò×éѧЧ¹ûÅú×¢£¬ÃÅˮƽÉÏÁ½¸ö×éÖÐ×îÖ÷ÒªµÄ¾úΪÄâ¸Ë¾úÃÅ¡¢ºñ±Ú¾úÃÅ¡¢·ÅÏß¾úÃźͱäÐξú£¨Í¼3 A£©¡£NCDs×éÓëNA×éÏà±È£¬KiritimatiellaeotaºÍLentisphaeraeµÄƵÂÊÏÔÖøïÔÌ£¨Í¼3 B£©¡£NA×éºÍNCDs×éÖ®¼ä±£´æ¿ÉÊÓ²ìµÄ¾ÛÀàºÍÏÔ×ŵÄÇ÷ÊÆ£¬ÕâÅú×¢³¦µÀ΢ÉúÎïȺÌåµÄʧµ÷ÓëNCDsÓйأ¨Í¼3 E£©¡£ ͼ3 NA×éºÍNCDs×éµÄ³¦µÀ΢ÉúÎïȺÆÊÎö ËùÓÐÑùÆ·Öй²ÓÐ1039¸ö²Ù×÷·ÖÀ൥λ£¨Í¼4 A£©£¬ÆäÖÐ17¸öϸ¾úȺÔÚNA×éÖи»¼¯£¬10¸öϸ¾úȺ¸»¼¯ÔÚNCDs×é¡£±ðµÄ£¬LEfSe(linear discriminant analysis effective size)È·¶¨ÁË7ÖÖ¶ÔNCDsÌØÒìÐÔµÄÖÖˮƽµÄ¾ú£¨Í¼4 D£©¡£ÓëNA×éÏà±È£¬NCDs×éÖÐÓë·¼Ïã×å°±»ùËáÉúÎïºÏ³ÉÏà¹ØµÄ;¾¶ºÍÈýôÈËáÑ»·£¨TCA»·£©Ïà¹ØµÄͨ·µÈÏÔÖøïÔÌ£¨Í¼4 E£©¡£ ͼ4 ÖÖÊôˮƽµÄ³¦µÀ΢ÉúÎïÒÔ¼°Ïà¹ØµÄÕ¹Íû¾úÖÖµÄÂþÑÜ Í¨¹ýÂÑ°×ÖÊ×éѧ¹²·¢Ã÷3306¸öÂÑ°×ÖÊ£¬ÆäÖÐ2712¸ö¿ÉÁ¿»¯£¨Í¼5 A£©¡£ÕâЩÂÑ°×Öʱ»¶¨Î»ÔÚϸ°ûÖÊ¡¢Íâϸ°û¿Õ¼ä¡¢ºËºÍÏßÁ£ÌåµÈϸ°ûÇøÓòÖУ¨Í¼5 B£©¡£Æ«×îС¶þ³Ë·¨(partial least squares discriminant analyis, PLS-DA)ÆÊÎöÅú×¢£¬NCDs×éºÍNA×é¼äÓÐÏÔ×ŵÄÇø·Ö£¨Í¼5 C£©¡£»ðɽͼÅú×¢£¬ÉÏе÷ϵ÷µÄ²î±ðÂÑ°×»®·ÖÓÐ57ºÍ30¸ö£¨ºìÉ«ÌåÏÖÏÔÖøÉϵ÷£¬À¶É«ÌåÏÖÏÔÖøϵ÷£©¡£Í¨¹ý´úл¹¦Ð§ÆÊÎö£¬·¢Ã÷²î±ð±í´ïµÄÂÑ°×ÖÊÖ÷Òª¼ÓÈë°±»ùËáÔËÊäºÍ´úл£¬ÒÔ¼°ÄÜÁ¿Éú²úºÍת»¯£¬ÕâÓëºê»ùÒò×éѧÆÊÎöµÄÕ¹ÍûЧ¹ûÒ»Ö£¨Í¼5 E£©¡£±ðµÄ£¬KEGGͨ·¸»¼¯ÆÊÎöҲ֤ʵÁËÕâЩ·¢Ã÷¡£Ïµ÷µÄÄòÒºÍâÃÚÌåÂÑ°×µÄ;¾¶¸»¼¯ÆÊÎöÏÔʾ£¬NCDs×éµÄÀÒ°±Ëá´úл¡¢É«°±Ëá´úлºÍ±ûͪËá´úлïÔÌ£¨Í¼5 F£©¡£Ïà·´£¬ÓëÑ×Ö¢ÐÔ³¦²¡¡¢°òë×°©¡¢IÐÍÌÇÄò²¡ºÍºã¾ÃÒÖÓôÖ¢µÈ¼²²¡Ïà¹ØµÄ;¾¶ÔÚ·ÇѬȾÐÔ¼²²¡×éµÄÉϵ÷ÂÑ°×Öи»¼¯£¨Í¼5 G£©¡£ ͼ5 ÄòҺĤϸ°ûµÄ¶¨Á¿ÂÑ°××éѧÆÊÎö »ùÓÚ·Ç°ÐÏò´úл×éѧÆÊÎöЧ¹û·¢Ã÷£¬ÕýÀë×Ӻ͸ºÀë×ÓģʽÏ´úлÎïµÄPLS-DAÆÊÎö¾ùÅú×¢NA×éºÍNCDs×é¼ä±£´æÏÔ×ŵÄÇø·Ö£¨Í¼6 A-B£©¡£ÕýÀë×ÓģʽÖУ¬ÓëNA×éÏà±È£¬¾ùÓÐ54¸öÉϵ÷¼°Ïµ÷µÄ²î±ð´úлÎ¶ø¸ºÀë×ÓģʽÖУ¬ÓÐ12¸öºÍ63¸ö´úлÎïÏÔÖøÉÏе÷ϵ÷£¨Í¼6 C£©¡£Ñ¡È¡ÁËÇ°20¸ö²î±ð´úлÎï¹¹½¨ÈÈͼ£¨Í¼6 D£©£¬½øÒ»²½Æ¾Ö¤±äÁ¿Í¶Ó°Ö÷ÒªÐÔ(Variable importance in the projection, VIP)Öµ¾ÙÐл®·Ö£¬¹¹½¨Á˿ɱäÖ÷ÒªÐÔͼ¡£ÔÚNCDs×éÖÐÊӲ쵽ÁËôÇ»ùÖ¬·¾ËáÖ¬·¾Ëáõ¥(fatty acid esters of hydroxy fatty acids, FAHFAs)ÏÔÖøïÔÌ£¨Í¼6 E£©¡£¶Ô²î±ð´úл²úÆ·µÄËùÓÐ;¾¶¾ÙÐÐKEGGͨ·ÆÊÎö£¬Ð§¹ûÅú×¢ÑÌËáºÍÑÌõ£°·´úл¡¢L-É«°±ËáÉúÎïºÏ³É¡¢·ºËáºÍ¸¨Ã¸AÉúÎïºÏ³É¡¢Ò¶ËáÉúÎïºÏ³É¡¢²»±¥ºÍÖ¬·¾ËáÉúÎïºÏ³É¡¢ÄûÃÊËáÑ»·¡¢±ûͪËá´úлºÍÁò°·ËØ´úлÓëNCDsÏà¹Ø£¨Í¼6 F£©¡£ ͼ6 NA×éºÍNCDs×éµÄ·à±ãÑù±¾µÄ´úл×éѧÆÊÎö ÐÐΪʵÑéЧ¹ûÅú×¢£¬FMT»áÓ°ÏìNCDs×éÄ£×ÓСÊóµÄ¿Õ¼äѧϰºÍÓ°Ïó£¨Í¼7£©¡£ ͼ7 ʹÓÃYÃÔ¹¬¡¢ÐÂÎïÌåʶ±ðºÍMorrisË®ÃÔ¹¬²âÊÔÆÀ¹ÀFMTÏÂСÊó¿Õ¼äѧϰºÍÓ°ÏóÌåÏÖÄÜÁ¦ ΪÁ˽øÒ»²½ÑéÖ¤´úл;¾¶£¬Ñо¿Ö°Ô±Í¨¹ýÍøÂçºÍÆÊÎö²î±ðʱ¼äµãµÄ·à±ã¿ÅÁ££¬ÆÀ¹ÀÁËFMT¶Ô³¦µÀ΢ÉúÎïȺÌåµÄÓ°Ï죬ͨ¹ýѪÇåºÍº£ÂíÌåµÄ°ÐÏò´úл×éѧÆÊÎöÑéÖ¤·¢Ã÷£¬ÄûÃÊËá¡¢¸»ÂíËá¡¢çúçêËá¡¢ÑõÎì¶þËá¡¢±ûͪËáºÍÉ«°±ËáµÄÏà¶Ôº¬Á¿¶¼±¬·¢ÁËÏÔÖøת±ä£¬ÖµµÃ×¢ÖصÄÊÇÄûÃÊËáµÄÏà¶Ôº¬Á¿ÔÚѪÇåÖÐÏÔÖøÔöÌí£¬µ«ÔÚº£ÂíÖÐÏÔÖø½µµÍ£¨Í¼8£©¡£ ͼ8 FMTСÊ󳦵À¾úȺµÄºê»ùÒò×éÆÊÎöÒÔ¼°ÑªÇåºÍº£ÂíµÄ´úл×éѧÆÊÎöµÄÆÀ¹À Ñо¿Ö°Ô±Ê¹ÓÃÖ§³ÖÏòÁ¿»ú(support vector machine, SVM)½á¹¹Ä£×Ó£¬½«¸öÌå·ÖÀàΪNA×éºÍNCDs×飨ͼ9£©£¬Óúê»ùÒò×é¡¢´úл×éѧºÍEEGÊý¾ÝÈýÖÖ×éºÏÖлñµÃµÄÌØÕ÷À´ÑµÁ·Ä£×Ó¡£×îÖÕ£¬¶à×éѧҪÁìµÖ´ïÁË׼ȷÂÊΪ92.69%£¬×¼È·ÂÊΪ94.44%£¬ÕÙ»ØÂÊΪ89.47%£¬AUCֵΪ0.9641£¨Í¼9 G£©£¬Í¬Ê±·¢Ã÷Ëæ×Å×éѧȪԴµÄÔöÌí£¬Ê¹ÓÃÏàͬÊýÄ¿µÄÌØÕ÷£¬Ä£×ÓµÄÐÔÄÜÖð½¥Ìá¸ß¡£ ͼ9 ²î±ð×éѧϵĻúеѧϰÄÜÁ¦ ÔÚÕâÏîÑо¿ÖУ¬Õ¹ÏÖÁ˼¸¸öÒªº¦·¢Ã÷£ºÊ×ÏÈ£¬ÊӲ쵽PSD½µµÍ£»Æä´Î£¬È·¶¨ÁË·¼Ïã°±»ùËáÉúÎïºÏ³ÉºÍÈýôÈËáÑ»·ÖеÄ×ÌÈÅ¡£ÕâЩ×ÌÈÅÓëÌض¨Î¢ÉúÎïˮƽµÄÔöÌíÓйأ»±ðµÄ£¬·¢Ã÷Ìض¨´úл²úÆ·µÄˮƽ½µµÍ£¬ÕâЩ·¢Ã÷ÅäºÏÌṩÁ˶ԷÇѬȾÐÔ¼²²¡²¡ÀíÐÄÀíѧDZÔÚÌØÕ÷µÄ¿´·¨£»×îºó£¬»úеѧϰģ×ÓµÄÓ¦Óôó´óÌá¸ßÁ˼ÓÈëÕ߶ԷÇѬȾÐÔ¼²²¡µÄÇø·Öˮƽ¡£Òò´Ë£¬¶à×éѧÌØÕ÷ºÍÄÔµçͼÊý¾ÝµÄÈÚºÏΪ·ÇѬȾÐÔ¼²²¡»¼ÕߵĸöÐÔ»¯Õï¶ÏÌṩÁËÌرðµÄÔ¶¾°¡£