Á¤º¸°úÇÐȸ³í¹®Áö (Journal of KIISE)
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
EFA-DTI: Edge Feature AttentionÀ» È°¿ëÇÑ ¾à¹°-Ç¥Àû »óÈ£ÀÛ¿ë ¿¹Ãø |
¿µ¹®Á¦¸ñ(English Title) |
EFA-DTI: Prediction of Drug-Target Interactions Using Edge Feature Attention |
ÀúÀÚ(Author) |
¿¡¸£Çܹپ߸£ ÀÚ´ã¹Ù
±è¼öÇå
ÀÌÇö¼ö
±èÈÁ¾
Erkhembayar Jadamba
Sooheon Kim
Hyeonsu Lee
Hwajong Kim
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 48 NO. 07 PP. 0825 ~ 0834 (2021. 07) |
Çѱ۳»¿ë (Korean Abstract) |
½Å¾à°³¹ßÀº ÀǾà ÈÇÐ, ½Ã½ºÅÛ ¹× ±¸Á¶ »ý¹°ÇÐ, ´õ ³ª¾Æ°¡ ÀΰøÁö´É¿¡ À̸£±â±îÁö ´Ù¾çÇÑ Çй®À» ÇÊ¿ä·Î Çϱ⠶§¹®¿¡ ³À̵µ°¡ ³ôÀº ºÐ¾ß¶ó°í ÇÒ ¼ö ÀÖ´Ù. ƯÈ÷, ¾à¹°-Ç¥Àû »óÈ£ÀÛ¿ë(DTI) ¿¹ÃøÀº ¹æ´ëÇÑ ¾çÀÇ ÈÇÕ¹°·ÎºÎÅÍ Áúº´À» Ä¡·áÇÒ ¼ö ÀÖ´Â Èĺ¸ ¹°ÁúÀ» µµÃâÇس»´Â °úÁ¤À¸·Î, ½Å¾à °³¹ß °úÁ¤¿¡ ÀÖ¾î ÇÙ½É ¿ä¼Ò´Ù. ÃÖ±Ù¿¡´Â ÄÄÇ»ÅÍ ¼º´ÉÀÌ ºñ¾àÀûÀ¸·Î ¹ßÀüÇÔ¿¡ µû¶ó, DTI ¿¹Ãø¿¡ ¼Ò¿äµÇ´Â ¿©·¯ Ãø¸éÀÇ ºñ¿ëÀ» ÁÙÀÌ°íÀÚ ÀΰøÁö´É ½Å°æ¸ÁÀ» È°¿ëÇÏ´Â ¿¬±¸°¡ È°¹ßÈ÷ ÁøÇàµÇ°í ÀÖ´Ù. µû¶ó¼, º» ³í¹®¿¡¼´Â Edge Feature AttentionÀ» Àû¿ëÇÑ Graph Net Embedding ¹× Fingerprint¸¦ È°¿ëÇÑ ¾à¹° Ç¥Çö »ý¼º°ú ProtTrans¸¦ È°¿ëÇÑ ´Ü¹éÁú Ç¥Çö »ý¼ºÀ» ÅëÇØ ¾à¹°°ú Ç¥Àû ´Ü¹éÁú °£ÀÇ »óÈ£ÀÛ¿ë ¼öÄ¡¸¦ ¿¹ÃøÇÏ´Â ¸ðµ¨À» Á¦¾ÈÇÑ´Ù. ÇØ´ç ¸ðµ¨Àº ±âÁ¸ÀÇ DTI ¿¬±¸¿¡¼ °¡Àå ÁÁÀº ¼º´ÉÀ» º¸¿´´ø DeepDTA, GraphDTAº¸´Ù ³ôÀº ¼º´ÉÀ» ´Þ¼ºÇÏ¿´À¸¸ç, ÀÌ¿¡ ´ëÇÑ ½ÇÇè ¹× °á°ú¸¦ ±â¼úÇÏ¿´´Ù. |
¿µ¹®³»¿ë (English Abstract) |
Drug discovery is a high-level field of research requiring the coordination of disciplines ranging from medicinal chemistry, systems biology, structural biology, and increasingly, artificial intelligence. In particular, drug-target interaction (DTI) prediction is central to the process of screening for and optimizing candidate substances to treat disease from a nearly infinite set of compounds. Recently, as computer performance has developed dramatically, studies using artificial intelligence neural networks have been actively conducted to reduce the cost and increase the efficiency of DTI prediction. This paper proposes a model that predicts an interaction value between a given molecule and protein using a learned molecule representation via Edge Feature Attention-applied Graph Net Embedding with Fixed Fingerprints and a protein representation using pre-trained protein embeddings. The paper describes architectures, experimental methods, and findings. The model demonstrated higher performance than DeepDTA and GraphDTA, which had previously demonstrated the best performance in DTI studies. |
Å°¿öµå(Keyword) |
BERT
»çÀüÇнÀ¸ðµ¨
°¨Á¤ ºÐ¼®
¿ÜÀû °áÇÕ
Áö½Ä Áõ·ù
°¨Á¤ ÀÚÁú
BERT
pre-trained model
sentiment analysis
external fusing
knowledge distillation
sentiment features
ÀΰøÁö´É
±×·¡ÇÁ ÀÓº£µù
¾îÅÙ¼Ç ¸ÞÄ¿´ÏÁò
¾à¹°-Ç¥Àû »óÈ£ÀÛ¿ë
AI
graph embedding
attention mechanism
drug-target interaction
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|