Kristiina Ausmees

doctoral/PhD student at Department of Information Technology, Division of Scientific Computing

Email:
kristiina.ausmees[AT-sign]it.uu.se
Telephone:
+4618-471 2978
Visiting address:
Room POL 2404 ITC, Lägerhyddsvägen 2, hus 2
752 37 UPPSALA
Postal address:
Box 337
751 05 UPPSALA

Short presentation

I work in an interdisciplinary project where scientific computing is applied in the study of human evolutionary and demographic history. My main focus is on computational methods for genetic inference and their application to ancient human DNA; designing new methods and adapting existing ones to the context of highly sparse and uncertain data.

My interests include statistical data analysis, machine learning and cloud technologies.

My project is a collaboration with the Jakobsson Lab at EBC.

Keywords: scientific computing data science statistical modeling and machine learning next generation sequencing ancient dna

My courses

Biography

I graduated with a masters degree in Computer Science from Uppsala University in 2016 and started my PhD studies in Scientific Computing the same year.

Research

The goal of my PhD project is to develop computational methods for analysis of ancient DNA. The main focus is on statistical models adapted to handling high levels of sparsity and uncertainty in the data.

I've mainly worked on developing and evaluating methods for imputation, or the inference of unobserved genotypes. As this can be used to increase the information content of samples, it has the potential to increase the scientific returns of ancient data. The statistical methods that I work with include sequential state space models that explicitly model biological processes, as well as data-driven models such as neural networks.


A parallel but related track of my project is related to efficient handling of large data sets. In genomics, as in many other scientific disciplines, the large amounts of data that are analyzed put great demands on efficiency of compute and storage solutions. As access to high-performance computing facilities is not always available, the use of cloud-based and distributed services has seen an increase. An example of such an application that we have developed is the BAM Search Infrastructure (BAMSI), a cloud-based framework for distributed filtering of genomic data sets with the goal of facilitating storage and data management of scientific data sets.

Some of my projects are available on GitHub.

Publications

Please contact the directory administrator for the organization (department or similar) to correct possible errors in the information.