Memory-efficient genotype storage

Razvojem novih i poboljšanih metoda sekvenciranja, te opadanjem cijene sekvenciranja, dolazi se do sve većih skupova genomskih podataka i to sve većom brzinom. Sukladno tome, javlja se problematika memorijski učinkovitog skladištenja takvih podataka. Razvijene su razne metode za sažeti prikaz genotipova, od posebnih formata zapisa do algoritama za sažimanje. Ipak, nijedna od dosadašnjih tehnika nije se pokazala dugoročno poželjnom. GenoType Compressor (GTC) novi je algoritam specijaliziran za sažimanje podataka o genotipovima zapisanim u „variant call formatu” (VCF). Prva istraživanja pokazuju njegovu premoć u stupnju kompresije, ali i u vremenu pristupa sažetim podacima nad sličnim algoritmima. U ovom radu implementirana je osnovna verzija algoritma GTC te su analizirane performanse algoritma s obzirom na različite ulazne skupove i parametre. Također, usporedio sam implementaciju ostvarenu u okviru ovog rada s originalnom implementacijom GTC-a.With the development of new and improved sequencing methods and with the reducing costs of genotype sequencing, more and more data is being collected with increased rates. This leads us to a problem of managing memory-efficient storage of such data. Various different methods of compressed data representation have been developed ranging from special data formats to compression algorithms. However, neither of former methods proved to be sufficiently acceptable in the long term. GenoType Compressor (GTC) is a new algorithm highly specialised for compression of genotype data stored in „variant call format” (VCF). Preliminary research place GTC high above it's competition in terms of compression rate and lower data access times. In this paper, a basic version of GTC algorithm is implemented and its performance is analyzed on different input data sets. Moreover, the implementiation of GTC developed in this thesis is compared with the original implementation of GTC algorithm