Nobody knows the origin of most of the DNA from archeological samples. The analysis of DNA material gathered from archeological specimens (e.g., from teeth and bones) has made possible to give better answers to these questions. This ancient DNA (aDNA) can provide information about ancestral genetic characteristics, population dynamics, migrations and specific cultural traits as reflected, for example, in the burial rituals. Modern sequencing technologies allow us to obtain vast amounts of DNA data at low cost. The scientific challenge is to identify the origin of each DNA sequence and correlate them with the evolutionary history of the specimen. Usually, an important part of the aDNA sequences cannot be matched to the host species. These sequences may come from the microbiome inside the host or the surrounding environment. A useful approach for this problem is to use metagenomics tools based on the sequence composition. In contrast to alignment based tools, these composition methods do not require a close reference genome. Instead they rely on the statistical properties of DNA sequences that tend to be conserved between close related taxa. In this project we aim to improve the current methods for composition-based taxonomical classification by adapting state-of-the-art metagenomic tools and developing new ones specifically tailored to the characteristics of ancient DNA.