top of page
長方形
Life Science Knowledge Bank
Drug Discovery System
AI-KS-LSKB-logo_透明.png
Use Bioinformatics Data to Get a 360-Degree View of Your Development
LSKB has been updated to Ver7, including features of MoA (mechanism of action) created from deep learning and original ontology, functions of searching chemicals and targets, an automated process of workflow, and a decision making support of ElpisMap.
About LSKB
LSKB Introduction

LSKB Introduction

Play Video
From Disease
Finding targets or chemicals  with various methods from a disease
Searching targets and chemicals based on a mechanism and function

By combining original MoA, which was extracted from AI, a new searching method, based on a mechanism,  has been developed.

Find new targets and new drugs from disease

Find new targets and new drugs from disease

Play Video
Categorized information of disease targets
  • By classifying targets with ontology classes, they can be efficiently narrowed down.

  • This enables access to activity values along with target details and directly connected assay information.

disease2targetclass.png
From Target (Gene or Protein)
タンパク質や遺伝子の周辺情報を利用して効率よくターゲットを探索
ターゲットから治療薬や化合物を探す

薬剤やアッセイ経由の化合物をアッセイ単位、Single Protein Assay, Activityでグループ化して見ることができます。薬剤では、Max Phase, Withdrawn, ATCとともにまとめています。

関連化合物の調査は遺伝子から行いますが、関連するであろう化合物を広く探し、化合物単位でアッセイやGene Expression, MoA, Withdrawn, ATC, 文献などの情報を一度に見ることができます。

ターゲットから疾患を探す

​疾患を各Phase毎にグループ化できるので、ターゲットに対して疾患候補を効率的に

target2disease.png
From Target (Gene or Protein)
Searching targets by using surrounding information of proteins or genes.
Searching drugs and chemicals from a target

You can see grouped drugs or chemicals via assay with units of assay, Single Protein Assay, or Activity.

For drugs, they are categorized with Max Phase, Withdrawn, and ATC. 

Examination of related chemicals can start from a gene. By searching related chemicals widely, you can see assay, Gene Expression, MoA, Withdrawn, ATC, and publication with a chemical annotation.

Searching diseases from a target

Since diseases can be categorized by each phase, disease candidates against a target can be efficiently searched.  

target2disease.png
Referring evaluated similar proteins with three methods

Results of Binding site, CD-HIT, Blast, RPS-Blast, and DELTA-Blast are contained. 

similar_protein.png
From Chemical
Searching surrounding information based on chemical annotations and mechanisms.
Obtaining pharmacology, MoA, metabolism enzymes, and similar chemical information.

For pharmacology, as MoA related information, you can see drugs or diseases, having the same mechanisms or MoA related metabolism enzymes.

For related chemicals, you can see metabolisms, chemicals with salts and desalinated chemicals.

Referring public Gene Expression data by utilizing chemicals

You can see terminologies, such as diseases, tissues, chemicals with title descriptions, the number of differentially expressed genes, and publications.

Results of a relationship between a chemical and a disease

You can see a relationship between a chemical and a disease, from GO, MoA, Max Phase, Clinical Trial, and publications in a matrix form.

imatinib_disease.png
Assay results summary by chemicals

Assay, Activity, ADME, and Toxicity will be displayed along with targets. Also, using the filtering function with an activity value or an assay time, you can search publications of your interest.

Target prediction using a chemical structure

 You can predict target and the activity based on similarity.

WorkFlow Tool

ワークフローツールは、多面的な要素による絞り込みなど複雑な検索を簡単に構築でき、かつ再利用できます。

データベースの検索、検索結果のフィルターなどを組み合わせることで、目的に合わせた検索やデータマイニングが行えます。

WorkFlow Tool
A searching tool, maximizing LSKB database
Obtaining related information from a target or a chemical

The workflow tool makes complex searching, such as filtering with multiple factors, easy and reuses the search history. 

By combining database search and filters of search results, you can execute your own searching or data-mining.

Annotation_Finder_Menu01.png
Preparation of a configured workflow

There is a configured workflow as an example. You can simply use it or modify it in your way.

Predifined_Workflow.png
How to use the workflow

This is a video introduction on how to use the system.

Workflow tool

Workflow tool

Play Video
Elpis Map
HTSデータの視覚化により、GO, No-Go, No-Needを判断しやすくする意思決定支援ツール
分子量の変化に伴う活性の変化の関係を視覚化

ElpisMapは、化合物の散布図の上に、例えば 分子量で3分割し、活性値などの別の特徴量で各グループの重心とその軌跡を表示できます。これにより入力した集合が 分子量の増加に伴い 活性の増強をしているか容易に判断できる 意思決定サポートツールです。

ヒトPIK3CGのアッセイにおいて 共通部分構造をもつ化合物の活性情報例

 ヒトPIK3CGのアッセイにおいて 共通部分構造をもつ化合物の活性情報について X軸に 分子量とY軸に 活性値でプロットしています。グレーの○は 分子量300未満、300以上350未満、350以上でカテゴリーをわけてその重心をプロットし、軌跡を矢印で示しています。

これによりこの骨格をもつ化合物が分子量の増加により ヒトPIK3CGに対する活性値が向上することが分かります。

Subst_kinase_02.png
ElpisMap_PK3CG_HUMAN_Subst_kinase_02a.png

先の共通部分構造をもつ化合物群のアッセイ情報から各PIK3CA、PIK3CB、PIK3CG、PIK3CDの活性値でプロットしています。下段の PIK3CG、PIK3CDに対する結果は、グレーの○は 分子量300未満、300以上350未満、350以上でカテゴリーに分割。その重心の軌跡からは 分子量増大において 活性が上昇し、この骨格の化合物セットは これらターゲットに有効であると示唆される。一方 上段のPIK3CA、PIK3CBに対する結果は、分子量増大において 活性が低下しており、更なる構造展開で活性の向上が困難であることを示唆しています。

ElpisMap_PIK3CA_PIK3CB_PK3CG_PK3CD_Subst_kinase.png
化合物情報のテーブル表示

PIK3CD の プロットにおいて、赤で囲んだ 活性値の高いサークルを選択することで、他の3つのプロットでも 選択されていることがわかる。さらにSelected Itemでにおいて選択した化合物情報がテーブルで表示されています。

ElpisMap_PIK3CA_PIK3CB_PK3CG_PK3CD_Subst_kinase_03.png
Elpis Map
Decision making support tool of GO, No-Go or No-Need from visualization of HTS data.
Visualization of change in activity along with change in molecular amount.

ElpisMap shows features of a divided chemical compound on a scatter graph. For example, it divides the compound in 3 ways with the molecule amount and displays the center of the balance for each group and its orbit with the feature amount, such as its activity value. This is a support tool to easily find out whether entered items of chemicals show the reinforcement of activity along with the increased molecule amount or better profiles could be expected.

Example: Activity information on chemicals, having common structures on human PIK3CG assay.

On an assay of human PIK3CG, activity values are plotted, using the x axis as chemical information, having common structural parts and the y axis as a molecule amount. The grey ○ plots centers by categorizing molecule amounts with less than 300, more than 300 and less than 350, and more than 350, and arrows indicate its trajectory. From this, you can see an activity value against the human PIK3CG is increasing from the increased molecule amount of the chemical with this framework.

Subst_kinase_02.png
ElpisMap_PK3CG_HUMAN_Subst_kinase_02a.png

From assay information of a chemical compound with a common structure, PIK3CA, PIK3CB, PIK3CG, and PIK3CD are plotted with their activity values. For the lower PIK3CG and PIK3CD, the grey ○ separates molecule amounts as categories with less than 300, more than 300 and less than 350, and more than 350, and looking at their tracing centers, their activities have increased based on the increased molecule amount, and these chemical sets of framework indicate that they are valid for targets. On the other hand, for the upper PIK3CA and PIK3CB, their activities have decreased, and this indicates that their activities may not improve to develop structures further.

ElpisMap_PIK3CA_PIK3CB_PK3CG_PK3CD_Subst_kinase.png
Display chemical information

On the plot of PIK3CD, by selecting a circle of the high activity value with the red color, you can see that other 3 plots are also selected. This information can be obtained from the table of Selected Item. 

ElpisMap_PIK3CA_PIK3CB_PK3CG_PK3CD_Subst_kinase_03.png
Saving calculation results

Calculated results, including used conditions, can be saved in an explorer format. Also, you can export the results in a dedicated format and share them with others. 

ElpisMap_sample.png
bottom of page