Establishment of an in silico phototoxicity prediction method by combining descriptors related to photo-absorption and photo-reaction
-
- Haranosono Yu
- Senju Pharmaceutical Co. Ltd., Toxicology Research Laboratory
-
- Kurata Masaaki
- Senju Pharmaceutical Co. Ltd., Toxicology Research Laboratory
-
- Sakaki Hideyuki
- Senju Pharmaceutical Co. Ltd., Toxicology Research Laboratory
書誌事項
- タイトル別名
-
- Establishment of an <i>in silico</i> phototoxicity prediction method by combining descriptors related to photo-absorption and photo-reaction
この論文をさがす
抄録
One of the mechanisms of phototoxicity is photo-reaction, such as reactive oxygen species (ROS) generation following photo-absorption. We focused on ROS generation and photo-absorption as key-steps, because these key-steps are able to be described by photochemical properties, and these properties are dependent on chemical structure. Photo-reactivity of a compound is described by HOMOLUMO Gap (HLG), generally. Herein, we showed that HLG can be used as a descriptor of the generation of reactive oxygen species. Moreover, the maximum-conjugated π electron number (PENMC), which we found as a descriptor of photo-absorption, could also predict in vitro phototoxicity. Each descriptor could predict in vitro phototoxicity with 70.0% concordance, but there was un-predicted area found (gray zone). Interestingly, some compounds in each gray zone were not common, indicating that the combination of two descriptors could improve prediction potential. We reset the cut-off lines to define positive zone, negative zone and gray zone for each descriptor. Thereby we overlapped HLG and PENMC in a graph, and divided the total area to nine zones with cut-off lines of each descriptor. The rules to prediction were decided to achieve the best concordance, and the concordances were improved up to 82.8% for selfvalidation, 81.6% for cross-validation. We found common properties among false positive or negative compounds, photo-reactive structure and photo-allergenic, respectively. In addition, our method could be adapted to compounds rich in structural diversity using only chemical structure without any statistical analysis and complicated calculation.
収録刊行物
-
- The Journal of Toxicological Sciences
-
The Journal of Toxicological Sciences 39 (4), 655-664, 2014
一般社団法人 日本毒性学会
- Tweet
詳細情報 詳細情報について
-
- CRID
- 1390001204903595008
-
- NII論文ID
- 130004904013
-
- NII書誌ID
- AN00002808
-
- COI
- 1:STN:280:DC%2BC2cbls1Gqtw%3D%3D
-
- ISSN
- 18803989
- 03881350
-
- NDL書誌ID
- 025758835
-
- PubMed
- 25056790
-
- 本文言語コード
- en
-
- データソース種別
-
- JaLC
- NDL
- Crossref
- PubMed
- CiNii Articles
-
- 抄録ライセンスフラグ
- 使用不可