Natural language processing of gene descriptions for overrepresentation analysis with GeneTEA

Data and models supporting "Natural language processing of gene descriptions for overrepresentation analysis with GeneTEA" (Boyle et al. 2025)File descriptions:GeneTEA.pkl, GeneTEA-yeast.pkl, PharmaTEA.pkl - pickled GeneTEA modelsFig2b/d.csv: Top terms in Figure 2 b/d.gProfiler_hsapiens_7-16-2025_4-35-07 PM__intersections.csv: g:GOSt results for Fig2b, downloaded from the g:Profiler site.gProfiler_hsapiens_7-16-2025_4-41-03 PM__intersections.csv: g:GOSt results for Fig2d, downloaded from the g:Profiler site.enrichr_sets_03_01_2025.csv: Enrichr database downloaded 3/1/2025, used for Figure 3 and S1.gene sets for connexin.gmt: Enrichr gene sets containing the term "connexin", downloaded from the Enrichr site.false_discoveries.csv: Benchmarking results for false discovery control in Figures 4 and S1.EF_hand_example.csv: Top terms and MedCPT scores for EF-hand example in Figure 4.[Hallmark or Experimentally Derived Queries]_scores.csv: Benchmarking results for [Hallmark or Experimentally Derived Queries] across joined top terms in Figures 4 and S2. The "joined_ranking" column corresponds to the MedCPT Relevance score across the top terms and "num_high_redundancy" contains the number of redundant term pairs.[Hallmark or Experimentally Derived Queries]_indiv.csv: Benchmarking results for [Hallmark or Experimentally Derived Queries] for each top term in Figures 4 and S2. The "indiv_ranking" column corresponds to the MedCPT Relevance score for a single term.Fig4_left/right: Examples of top terms shown in what is now Figure 5.gProfiler_hsapiens_3-13-2025_9-14-59 AM__intersections.csv: g:GOSt results for Fig4_left, downloaded from the g:Profiler site.gProfiler_hsapiens_2-11-2025_10-18-19 AM__intersections.csv: g:GOSt results for Fig4_right, downloaded from the g:Profiler site.

支撑《基于GeneTEA开展基因描述的自然语言处理以实现富集分析》（Boyle等，2025）研究的数据与模型
文件说明：
1. GeneTEA.pkl、GeneTEA-yeast.pkl、PharmaTEA.pkl：经pickle序列化存储的GeneTEA模型
2. Fig2b/d.csv：图2b与图2d中的排名靠前富集项
3. gProfiler_hsapiens_7-16-2025_4-35-07 PM__intersections.csv：对应图2b的g:GOSt分析结果，从g:Profiler官网下载所得
4. gProfiler_hsapiens_7-16-2025_4-41-03 PM__intersections.csv：对应图2d的g:GOSt分析结果，从g:Profiler官网下载所得
5. enrichr_sets_03_01_2025.csv：2025年3月1日下载的Enrichr数据库，用于图3及补充图S1
6. gene sets for connexin.gmt：从Enrichr官网下载的包含"connexin（连接蛋白）"术语的Enrichr基因集
7. false_discoveries.csv：图4及补充图S1中错误发现率控制的基准测试结果
8. EF_hand_example.csv：图4中EF-hand（EF手型结构）示例的排名靠前富集项及MedCPT得分
9. [Hallmark or Experimentally Derived Queries]_scores.csv：针对图4及补充图S2中合并排名靠前富集项的[Hallmark或实验衍生查询]的基准测试结果。其中"joined_ranking"列对应各排名靠前富集项的MedCPT相关性得分，"num_high_redundancy"列统计冗余项对的数量
10. [Hallmark or Experimentally Derived Queries]_indiv.csv：针对图4及补充图S2中各单个排名靠前富集项的[Hallmark或实验衍生查询]的基准测试结果。其中"indiv_ranking"列对应单个富集项的MedCPT相关性得分
11. Fig4_left/right：对应当前图5的排名靠前富集项示例
12. gProfiler_hsapiens_3-13-2025_9-14-59 AM__intersections.csv：对应Fig4_left（即当前图5左半部分）的g:GOSt分析结果，从g:Profiler官网下载所得
13. gProfiler_hsapiens_2-11-2025_10-18-19 AM__intersections.csv：对应Fig4_right（即当前图5右半部分）的g:GOSt分析结果，从g:Profiler官网下载所得