Nullomers are short sequences of DNA that do not occur in the genome of a species (for example, humans), even though they are theoretically possible. Nullomers must be under selective pressure - for example, they may be toxic to the cell. Some nullomers have been shown to be useful to treat leukemia, breast, and prostate cancer. They are not useful in healthy cells because normal cells adapt and become immune to them. Nullomers are also being developed for use as DNA tags to prevent cross contamination when analyzing crime scene material.
Nullomers are short sequences of DNA that do not occur in the genome of a species (for example, humans), even though they are theoretically possible. Nullomers must be under selective pressure - for example, they may be toxic to the cell. Some nullomers have been shown to be useful to treat leukemia, breast, and prostate cancer. They are not useful in healthy cells because normal cells adapt and become immune to them. Nullomers are also being developed for use as DNA tags to prevent cross contamination when analyzing crime scene material.
==Background== Nullomers are naturally occurring but potentially unused sequences of DNA. Determining these "forbidden" sequences can improve the understanding of the basic rules that govern sequence evolution. Sequencing entire genomes has shown that there is a high level of non-uniformity in genomic sequences. When a codon is artificially substituted with a synonymous codon, it often results in a lethal change and cell death. This is believed to be due to ribosomal stalling and early termination of protein synthesis. For example, both AGA and CGA code for arginine in bacteria; however, bacteria almost never use AGA, and when substituted it proves lethal. Such codon biases have been observed in all species, and are examples of constraints on sequence evolution. Other sequences may have selective pressure; for example, GG-rich sequences are used as sacrificial sinks for oxidative damage because oxidizing agents are attracted to regions with GG-rich sequences and then induce strand breakage. Moreover, it has been shown that statistically significant nullomers (i.e. absent short sequences which are highly expected to exist) in virus genomes are restriction recognition sites indicating that viruses have probably got rid of these motifs to facilitate invasion of bacterial hosts. Nullomers Database provides a comprehensive collection of minimal absent sequences from hundreds of species and viruses as well as the human and mouse proteomes.
Discovered by embedding cosine similarity (sentence-transformers MiniLM, 384-dim).