Benchmark of pre-processing pipelines of single cell RNA sequencing data
Single cell RNA sequencing (scRNA-Seq) is a very powerful tool to study the transcriptome. Several tools are available to pre-process the data it produces. However, no standard exists. The objective of this work is to compare four scRNA-Seq pre-processing tools and extract conclusions regarding the...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Fecha de publicación: | 2023 |
| País: | España |
| Institución: | Universitat Oberta de Catalunya (UOC) |
| Repositorio: | O2, repositorio institucional de la UOC |
| OAI Identifier: | oai:openaccess.uoc.edu:10609/148591 |
| Acceso en línea: | http://hdl.handle.net/10609/148591 |
| Access Level: | acceso abierto |
| Palabra clave: | scRNA-seq bioinformatics preprocessing Bioinformatics -- TFM Bioinformàtica -- TFM |
| Sumario: | Single cell RNA sequencing (scRNA-Seq) is a very powerful tool to study the transcriptome. Several tools are available to pre-process the data it produces. However, no standard exists. The objective of this work is to compare four scRNA-Seq pre-processing tools and extract conclusions regarding the advancements over recent years, and to determine the superior tool. The tools compared were (i) UMI Tools, one of the first published tools, which served as reference point, (ii) Salmon Alevin, a tool with many interesting features, (iii) Kallisto Bustools, a tool focused on computational efficiency and (iv) STARSolo, a recent implementation to a popular aligner. I designed 4 pipelines in bash to implement each tool, and I implemented a downstream analysis to evaluate the biological significance of the results. I compared the computational speed and efficiency, the count matrices produced and the biological results. For the comparisons, I used the datasets provided by Tabula Muris as ground truth. I found that Kallisto Bustools was significantly more efficient and faster, while UMI Tools was the slowest tool. The count matrices produced were consistent with the ground truth for UMI Tools and Kallisto Bustools while Salmon Alevin and STARSolo presented inconsistencies. The biological results were coherent, although Salmon Alevin showed problematics. I concluded that scRNA-Seq has progressed in recent years but more so in computational efficiency. Kallisto Bustools was the fastest and most consistent tool among those evaluated. |
|---|