Browsing by Subject "bioinformatics"

In recent decades, ancient DNA recovered from old and degraded samples, such as bones and fossils, has presented novel prospects in the fields of genetics, archaeology and anthropology. In Finland, ancient DNA research is constrained by the poor preservation of bones: they are quickly degraded by acidic soils, limiting the age of DNA that can be recovered from physical remains. However, some soil components can bind DNA and thus protect the molecules from degradation. Ancient DNA from soils and sediments has previously been used to reconstruct paleoenvironments, to study ancient parasites and diet and to demonstrate the presence of a species at a given site, even when there are no visible fossils present. In this pilot study, I explored the potential of archaeological sediments as an alternative source of ancient human DNA. I collected sediment samples from five Finnish Neolithic Stone Age (6,000–4,000 years ago) settlement sites, located in woodland. In addition, I analysed a lakebed sample from a submerged Mesolithic (10,000–7,000 years ago) settlement site, and a soil sample from an Iron Age burial with bones present to compare DNA yields between the two materials. Soil samples were converted into Illumina sequencing libraries and enriched for human mtDNA. I analysed the sequencing data with a customised metagenomics-based bioinformatic analysis workflow. I also tested program performance with simulated data. The results suggested that human DNA preservation in Finnish archaeological sediments may be poor or very localised. I detected small amounts of human mtDNA in three Stone Age woodland settlement sites and a submerged Mesolithic settlement site. One Stone Age sample exhibited terminal damage patterns suggestive of DNA decay, but the time of deposition is difficult to estimate. Interestingly, no human DNA was recovered from the Iron Age burial soil, suggesting that body decomposition may not serve as a significant source of sedimentary ancient DNA. Additional complications may arise from the high inhibitor content of the soil and the abundance of microbial and other non-human DNA present in environmental samples. In the future, a more refined sampling approach, such as targeting microscopic bone fragments, could be a strategy worth trialling.

Uterine leiomyomas are benign smooth muscle tumors arising in myometrium. They are very common, and the incidence in women is up to 70% by the age of 50. Usually, leiomyomas are asymptomatic, but some patients suffer from various symptoms, including abnormal uterine bleeding, pelvic pain, urinary frequency, and constipation. Uterine leiomyomas may also cause subfertility. Genetic alterations in the known driver genes MED12, HMGA2, FH, and COL4A5-6 account for about 90 % of all leiomyomas. These initiator mutations result in distinct molecular subtypes of leiomyomas. The majority of whole-genome sequencing (WGS) studies analyzing chromosomal rearrangements have been performed using fresh frozen tissues. One aim of this study was to examine the feasibility of detecting chromosomal rearrangements from WGS data of formalin-fixed paraffin embedded (FFPE) tissue samples. Previous results from 3’RNA-sequencing data revealed a subset of uterine leiomyoma samples that displayed similar gene expression patterns with HMGA2-positive leiomyomas but were previously classified as HMGA2-negative by immunohistochemistry. According to 3’RNA-sequencing, all these tumors overexpressed PLAG1, and some of them overexpressed HMGA2 or HMGA1. Thus, the second aim of this study was to identify driver mutations in these leiomyoma samples using WGS. In this study, WGS was performed for 16 leiomyoma and 4 normal myometrium FFPE samples. The following bioinformatic tools were used to detect somatic alterations at multiple levels: Delly for chromosomal rearrangements, CNVkit for copy-number alterations, and Mutect for point mutations and small insertions and deletions. Sanger sequencing was used to validate findings. The quality of WGS data obtained from FFPE samples was sufficient for detecting chromosomal rearrangements, although the number of calls were quite high. We identified recurrent chromosomal rearrangements affecting HMGA2, HMGA1, and PLAG1, mutually exclusively. One sample did not harbor any of these rearrangements, but a deletion in COL4A5-6 was found. Biallelic loss of DEPDC5 was seen in one sample with an HMGA2 rearrangement and in another sample with an HMGA1 rearrangement. HMGA2 and HMGA1 encode architectural chromatin proteins regulating several transcription factors. It is well-known that HMGA2 upregulates PLAG1 expression. The structure and functionality of HMGA2 and HMGA1 are very similar and conserved, so it might be that HMGA1 may also regulate PLAG1 expression. The results of this study suggest that HMGA2 and HMGA1 drive tumorigenesis by regulating PLAG1, and thus, PLAG1 rearrangements resulting in PLAG1 overexpression can also drive tumorigenesis. A few samples, previously classified as HMGA2-negative by immunohistochemistry, revealed to harbor HMGA2 rearrangements, suggesting that the proportion of HMGA2-positive leiomyomas might be underestimated in previous studies using immunohistochemistry. Only one study has previously reported biallelic inactivation of DEPDC5 in leiomyomas, and the results of this study support the idea that biallelic loss of DEPDC5 is a secondary driver event in uterine leiomyomas.