A complete summary of the notes written while reading the ScRAT paper and running its code.
[ScRAT] Paper
https://academic.oup.com/bioinformatics/article/40/2/btae067/7613064
[ScRAT] Paper Review
[ScRAT] Phenotype prediction from single-cell RNA-seq data using attention-based neural networks
[ScRAT] Phenotype prediction from single-cell RNA-seq data using attention-based neural networks
논문 https://academic.oup.com/bioinformatics/article/40/2/btae067/7613064정리 Attention 기반 신경망을 사용한 / 단일 세포 RNA-Seq 데이터의 / 표현형 예측Attention 기반으로 진행하는 것이 이 논문의 핵심!Attentino
doraemin.tistory.com
Code Execution
Customized dataset code execution
[ScRAT] customized dataset
Paper : Phenotype prediction from single-cell RNA-seq data using attention-based neural networks https://academic.oup.com/bioinformatics/article/40/2/btae067/7613064본 논문에서 언급된 ScRAT 방법으로 scRNA 분석하기[ScRAT] scRNA Analysis [ScR
doraemin.tistory.com
[ScRAT] customized dataset with CrossValidation
[ScRAT] customized dataset with CrossValidation
main.py에for loop 전체 변경하자.# 사전 생성된 split 파일을 기반으로 고정된 train/test 데이터셋으로 실험을 수행하기 위해 (2번코드) # => 주석처리# 3. for loop 직접 구성 (repeat × fold)for repeat in range(args.
doraemin.tistory.com
Results
[ScRAT] using SPILT Dataset
COVID 데이터셋에 대하여, 단일 데이터(.h5ad)가 아닌, 직접 (5개의 repeat) * (5개의 fold)를 나눠주고 지정하여 결과를 확인해보자먼저, 단일 데이터로 Hyperparameter Tuning을 해주었다HMAIL (HA) 논문의 baselin
doraemin.tistory.com
Data
[ScRAT] Exploratory Data (Summary)
Whether the dataset used for training scGPT and CellFM includes this dataset
[ScRAT Dataset] compare in cellxgene
[ScRAT Dataset] compare in CellFM
Workflow Summary
[ScRAT 흐름] Phenotype prediction from single-cell RNA-seq data using attention-based neural networks
[ScRAT] STEP 2. Attention layer
[ScRAT] sampling() function process
[ScRAT] using of cell type annotation
+ Mention of ScRAT in the HMIL paper
[Hierarchical MIL] compare AI Model
[Hierarchical MIL] compare AI Model
Hierarchical MIL (Multiple Instance Learning) 방법론과 비교하기 위해 사용된 최신 대표 모델에 대해 알아보자.(참고) Hier-MIML 논문 리뷰 : [Hierarchical MIL] Incorporating Hierarchical Information into Multiple Instance Learni
doraemin.tistory.com
'AI & Data Analysis > Deep Learning' 카테고리의 다른 글
| AMP + TF32 (2) | 2025.08.29 |
|---|---|
| [ScRAT] using of cell type annotation (0) | 2025.08.20 |
| [protocell4P] Customized Dataset (2) | 2025.08.14 |
| [Survey] Methods to Reduce Cell Count (2) | 2025.08.14 |
| [ScRAT] SPILT Dataset Results (4) | 2025.08.08 |
