Abstract
The field of Molecular Biology got started in earnest with the discovery of the molecular structure of DNA. This lead to a surge of interest into the relationships between DNA, RNA and proteins, and to the development of fundamental tools for manipulating those substances, such as cutting, ligating, amplifying, visualizing and size-selecting DNA. With these tools at hand it was possible to begin sequencing DNA, a process that took a leap forward in 2005 with the advent of Next Generation Sequencing (NGS). An inherent problem with NGS is that both the sequencing process and the library preparation introduce errors and biases. The massive amount of data generated by NGS, and the use of NGS in clinical settings, has created a demand for methods that can account for this and thereby making the sequencing data correct and reproducible.
Part 1 of this thesis briefly describes the development of Molecular Biology from the discovery of the molecular structure of DNA until today. Part 2 describes the development of error correcting and molecule counting methods, and part 3 describes the results of the papers of the thesis.
Paper I introduces the concept of Unique Molecular Identifiers (UMI) for reducing noise in molecular karyotyping and RNA sequencing data. Paper II compares the use of UMIs, a novel amplification-free method, and standard library preparation in Non-Invasive Prenatal Testing (NIPT) of fetal karyotype. Paper III uses the UMI concept together with single-cell tagged reverse transcription (STRT) to examine promoter preference in single cells. Finally paper IV use UMIs combined with PacBio sequencing to examine the full-length transcriptome in single cells.
