paper学习之unsupervised embedding of single-cell Hi-C data

---

Abstract

motivataion

1. significance of single cell Hi-C
2. problem in data process of Hi-C data: [However, Hi-C data analysis requires methods that take into account the unique characteristics of this type of data]
3. conclude the work of the paper: [In this work, we explore whether methods that have been developed previously for the analysis of bulk Hi-C data can be applied to scHi-C data. We apply methods designed for analysis of bulk Hi-C data to scHi-C data in conjunction with unsupervised embedding]

result

1. 得出结论HiCRep+MDS组合 outperform others
2. evidence： robust to extremely low per-cell sequencing depth；that this robustness is improved even further when high-coverage and low-coverage cells are projected together, and that the method can be used to jointly embed cells from multiple published datasets.

---

Introduction

1. 从单细胞水平上的High-throughput DNA sequencing technology发展的角度写，点明这些技术provide scientists with the opportunity to understand many aspects of fundamental functional processes in the cell, including gene regulation and DNA replication；最后指明一个fail to accurately capture the complexities of these types of data的分析方法是不行的。
2. 细化到单细胞引进的cell-to-cell variability分析问题，进一步点明FACS方法在state sorting上的局限性，需要methods for unsupervised analysis。
3. 列举single cell RNA-seq上的一些细胞差异性分析方法及其technique。
4. In this work，focus on single cell Hi-C data. The reasons: 1. sig cell Hi-C 的数据稀疏性；2. 该类数据的研究价值。
5. 介绍了单细胞Hi-C数据中研究异质性的一个特征量CDP的概念，及应用在哪些工作中。
6. 点名CDP这种特征量的局限性，忽略了了genome的结构化信息，不利于研究TAD，loop这些。
7. 引出bulk Hi-C中重复性分析的几种方法。
8. 提出paper的结论HiCRep+MDS结果较好，且robust to the number of contacts required per cell

---

Materials and Methods

数据集的基本信息

Similarity and distance measures for scHi-C contact maps

1. bulk Hi-C上提出的几种评估重复性方法的具体公式及similarity和distance的转换公式
2. cell orde具体角度计算
3. 排序结果的评估，提出了circular ROC指标。

Results

1. CROC的计算
2. 和 Nagano方法比较结果；并提出combined方法和结果验证；
3. few contacts per cell 鲁棒性实验；√
4. 从提高low coverage cell 排序准确性的角度提出 Joint projection with high-coverage cells improves phasing of low-coverage cells
5. Cell-cycle phased scHi-C data is indicative of
   replication timing 有点儿生物学×
6. Nagano和Flyamer联合分析。从cell cycle角度出发，只分析了G1 phase，侧面证明了所述方法的有效性，并针对Flyamer结果分布太离散给出解释：coverage或其他原因。