LETTER TO THE EDITOR The Mysterious Case of the Reverse Sequences

Abstract

I would like you to do a little experiment with me. Let’s suppose we get a bacterial SSU rRNA and we blast it at the NCBI [1]. The result we get is obvious: a large number of homologous sequences of many bacteria. Now let's do something else, we are going to reverse the same sequence. It is easy to do with the EMBOSS program Revseq [2]. The reverse sequence is written in the direction 3'-5'. For example 5’AGTC3’ becomes 3’CTGA5’. If we blast with the reverse sequence, what would we get? The answer is nothing. We should not get any significant result because the reverse sequence is meaningless, sequences are always written in the 5 '-3' direction in the databases. Well, have you already done the blast with the reverse sequence? You are surprised because there are significant results! How is it possible? By this method I have detected more than 300 SSU rRNA reversed sequences in the NCBI non-redundant nucleotide database (Table 1). I think researchers who sent the sequences were wrong and sent the sequences in 3'-5' direction, but why? One explanation is that people make mistakes. The rRNA sequences are commonly used in identification and phylogeny of species. It is unavoidable, If you do one thing many times, sooner or later you make a mistake. This may explain the abundance but not why people makes mistakes. I want to propose two explanations that are not mutually exclusive. The first one says that it depends on the language of the researcher who submitted the sequence to the database. Genetics was developed in western culture, where texts are read and written horizontally left-to-right (LFT), for that reason we write DNA sequences in the 5’-3’ direction, the 5’ end on the left, the 3’ end on the right. Some oriental languages are written right-to-left (RTL). Muslims and jews write horizontally in the RTL direction. The chinese and japanese write top-to-bottom (TTB) but they can both read the columns LFT or RTL. My guess is that researchers who mother-tongue is not LFT use RTL and/or TTB text editors that change the direction of the sequences. Table 1 shows the places of origin of the sequences, 57% of them belong to non-LFT countries. I have not included the sequences from