Wilhelmlab/prospect-ptms-charge
收藏资源简介:
PROSPECT PTMs数据集是一个用于质谱分析中前体电荷状态预测的机器学习数据集。该数据集由慕尼黑工业大学Wilhelmlab整理,基于Zenodo上的PROSPECT PTMs数据集。数据集包含两个配置:default和holdout,分别用于训练/验证/测试和最终模型评估。数据集的结构包括多个特征,如修改后的序列、原始文件、扫描号、包、最丰富的电荷状态、观察到的电荷状态和电荷状态分布。数据集不包含个人、敏感或私人数据。
The PROSPECT PTMs dataset is a mass-spectrometry dataset for applied machine learning in proteomics, specifically for precursor charge state prediction. Curated by Wilhelmlab at the Technical University of Munich, the dataset is based on the PROSPECT PTMs datasets hosted on Zenodo. It includes two configurations: default and holdout, intended for training/validation/testing and final model evaluation, respectively. The dataset structure comprises features such as modified sequence, raw file, scan number, package, most abundant charge state, observed charge states, and charge state distribution. The dataset does not contain any personal, sensitive, or private data.
数据集概述
数据集名称
- 名称: PROSPECT PTMs - Precursor Charge Prediction
数据集配置
- 配置名称: default, holdout
数据集特征
- 特征名称: modified_sequence, raw_file, scan_number, charge_by_max_intensity, one_hot_most_abundant_charge, charge_state_vector, normalized_intensity_distribution
- 数据类型: string, int64, float64
数据集分割
- default配置:
- train: 1135350 examples, 3123318666 bytes
- val: 326007 examples, 903602957 bytes
- test: 160572 examples, 453726869 bytes
- holdout配置:
- test: 42270 examples, 165662409 bytes
数据集大小
- default配置: 4480648492 bytes
- holdout配置: 165662409 bytes
下载大小
- default配置: 460990614 bytes
- holdout配置: 10112261 bytes
数据集用途
- 用途: 用于预测肽序列的前体电荷状态
数据集结构
- default配置: 基于所有Zenodo PROSPECT PTMs数据集
- holdout配置: 基于PROSPECT Test-PTM数据集
数据集加载示例
python
main data for training and evaluation
main_dataset = load_dataset("Wilhelmlab/prospect-ptms-charge")
holdout data for final benchmarking
holdout_dataset = load_dataset("Wilhelmlab/prospect-ptms-charge", "holdout")




