As the key oncogenic drivers in non-small cell lung cancer (NSCLC), various mutations of epidermal growth factor receptor (EGFR) with variable drug sensitivities have been the major obstacle for personalized medicine. A platform for drug sensitivity retrieval and reliable drug sensitivity prediction is desired for physician-scientists and clinicians. For the purpose, we built a database, namely D3EGFRdb, with the information of a total of 1,158 patients harboring EGFR mutations. Besides, we developed a deep learning model, namely D3EGFRAI, for drug sensitivity prediction for EGFR mutation-driven NSCLC. Accordingly, we constructed a web server, namely D3EGFR, to provide drug response retrieval through existing patient cases with specific clinical information and outcomes, and to predict drug response for patients harboring either common or rare or newly identified EGFR mutations. D3EGFR is free and open to all users and there is no login requirement.

This manual will help you become familiar with the basic operation procedures and precautions of D3EGFR as soon as possible.
download the manual

1. D3EGFRdb

D3EGFRdb is a clinical database about the characteristics and outcomes of EGFR-mutated lung cancer patients treated with EGFR-TKIs. At present, all of the clinical cases are from literature reports, which are annotated with a definite response of complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD) and other clinicopathologic characteristics. Users can query a certain mutation type to obtain the effects of existing drug therapy, as an important reference for drug recommendation.
Module 1: Drug Response Retrieval
• Input
Enter the mutant type of interest in the search box. Some examples of input mutation types are listed as follows:
♦ Wild Type (no mutation): WT
♦ Point mutation: L858R
♦ Deletion mutation: E746_A750del, V834del
♦ Insert mutation: D770_N771insSVD
♦ Duplicate mutation: A767dupASV
♦ Deletion-insertion mutation: L747_P753delinsS, D770delinsGY
♦ Complex mutation: E746_A750del+L858R

• Output
The clinical search results include each clinical response proportion of patients under different drug treatments. The greater the proportion of good responses (CR, PR), the better the drug's therapeutic effect against the mutant. The individual cases are listed below one by one, and the user can click on the reference to link to the original source.

Module 2: Clinical cases summary
All collected clinical cases are provided here.

2. D3EGFRAI

The clinical search results include each clinical response proportion of patients under different drug treatments. The greater the proportion of good responses (CR, PR), the better the drug's therapeutic effect against the mutant. The individual cases are listed below one by one, and the user can click on the reference to link to the original source.
Input and Output
Enter the mutant type of interest in the search box for predicting its drug sensitivity and drug response to EGFR-targeted drugs.

作为非小细胞肺癌的关键致癌驱动因子，表皮生长因子受体（EGFR）的众多突变由于其多变的药物敏感性问题，一直是精准医疗的主要障碍。也正因如此，临床科学家和临床医生亟需可用于药物敏感性检索和可靠预测药物敏感性的平台。为此，我们建立了包含1158名EGFR突变患者信息的D3EGFRdb数据库，并开发了可预测EGFR突变驱动的非小细胞肺癌药物敏感性的D3EGFRAI模型。在此基础上，我们构建了本网站平台，以通过现有突变患者病例提供药物反应检索，并预测携带常见或罕见或新发现的EGFR突变类型的患者的药物反应。所有用户可免费访问并使用D3EGFR网站，没有登录要求。

本手册将帮助您尽快熟悉D3EGFR的基本操作步骤和注意事项。下载手册

1. D3EGFRdb

D3EGFRdb是一个关于EGFR突变肺癌患者临床特征和治疗结果的临床案例数据库，收录的患者均接受EGFR酪氨酸激酶抑制剂作为治疗方案。目前，所有的临床病例均来自文献报道，并注明了明确的完全缓解（CR）、部分缓解（PR）、疾病稳定（SD）或疾病进展（PD）的反应以及其他临床病理特征。用户可以查询某种突变类型以获得现有药物对具有该突变型患者的治疗情况，作为选择药物的重要参考。
功能1：突变案例
在搜索框中输入感兴趣的突变类型。突变类型格式如下所示：
• 输入：
输入格式举例：
♦ 野生型：WT（无突变）
♦ 点突变：L858R（表示第858位的Leu被Arg取代）
♦ 缺失突变：E746_A750del（表示从第746位的Glu到第750位的Ala发生了缺失）
♦ 插入突变：D770_N771insSVD（表示从第770位的Asp到第771位的Asn之间插入了SerValAsp）
♦ 重复突变：A767_V769dupASV（表示从第767位的Ala到第769位的Val重复了一次）
♦ 插入缺失突变：L747_P753delinsS（表示从第747位的Leu到第753位的Pro缺失，同时被Ser取代）
♦ 复杂突变：E746_A750del + L858R（上述突变类型的组合）

• 输出：
突变案例检索结果显示了不同药物治疗下患者的各临床反应比例。其中，良好反应（CR、PR）的比例越大，药物对该突变型患者的治疗效果就越好。具体的单独患者案例在页面下方逐一列出，用户可以单击链接找到原始参考文献。

功能2: 案例汇总
此处提供了所有收集的临床病例。

2. D3EGFRAI

D3EGFRAI是一个可用于预测 EGFR 突变驱动的非小细胞肺癌的药物敏感性和药物反应的模块。用户可以通过查询某种突变类型来预测获批药物对该突变型可能的治疗效果。新突变型的计算时间在10秒左右。
输入和输出
在搜索框中输入感兴趣的突变体类型，以预测其对EGFR靶向药物的药物敏感性及药物反应。