Generally, it is easy to spot insertion sequences in bacterial genomes simply by sequence similarity. If you find a putative insertion sequence, go to IS Finder at https://www-is.biotoul.fr/index.html and do a BLAST search to see if it has already been registered. If not, please consider naming and submitting the novel insertion sequence to the IS Finder database.

You can use ICEberg (http://db-mml.sjtu.edu.cn/ICEberg/) to scan your genome for integrative and conjugative elements. For the identification of integrated prophage regions you can try PHAST (http://phast.wishartlab.com/).

The best approach is to scan your genomes for repetitive sequences (a quick approach is Nucmer). If you have repetitive regions you can classify/group them in different "families" (e.g. RECON). After that characterize them on ISSaga.

Thanks to all of you, I'm going to work with what you proposed. Best,

For those who are interested in this question, I found that review too: http://www.nature.com/hdy/journal/v104/n6/full/hdy2009165a.html

Best,

Question: How can I scan bacterial genomes to find known and unknown mobile elements?

To address the question, I will share my experiences which are mostly from sequence inspection. Generally, a genome of a bacterium contains multiple copies (including few deletion mutant sequences) of an IS of element, and also various kinds of ISs. A good start is by locating and collecting nucleotide sequences of transposases (or invertases), as well as their similar sequences found in the complete genome (using blast search). To search for structures, all the sequences as well as few hundred base pairs up and down stream should be included in the collection. It is important to employ the sequences with the same gene orientation (plus/plus or plus and complementary in case of plus/minus). After the sequences have been collected, try identifying typical IS’s characteristics that includes promoter-like sequences, inverted repeat sequences and/or duplicated direct repeat sequences (short oligonucleotide representing hall mark of transposition events). These sequences are usually located at up and down streams of the Tnps. Basically, multiple alignments of the nucleotide sequences (including deletion mutant sequences) are a powerful tool in grouping according to homologous sequences and identifying conserved terminal ends of many elements. Once the IS elements are classified based on their nucleotide similarity. Each categorized homologous copies (of an element) should be repeating the multiple alignments. This time is aim to identify the ends of the elements. Usually, the upstream of a bacterial tnp contain a promoter-like sequence (approx.100 bp-long) and a putative ribosome binding site (few bases away from initiation codon). In some cases, short direct repeat sequences (flanking the both termini) can be observed in some particular copies but many contain none. The direct repeats are short nucleotide sequences that were proved to be duplicated during some transposition events and thus it will be helpful to locate the inverted repeat ends. Once the terminal sequences of an element can be located, the left and right inverted sequences (IRs) will be easily matched (by observation) from reverse complementary sequences of the 3’-end to the 5’-end sequence. However, they are frequently imperfect match and some may not follow the rule. If two or more adjacent tnp genes are located in close proximity in the complete genomes, it is likely that they may share common IRs. Composite transposon is also possible found in some cases.

Best regards,

For the sequenced bacterial genomes or metagnomes, you can try ISEScan, https://github.com/xiezhq/ISEScan. It can output both complete inssertion sequence copies and incomplete insertion sequence copies. If you are only interested in the complete insertion sequence copies, you can change the option in its configure file (constants.py) to let ISEScan only output complete insertion sequence copies. I am the developer of ISEScan and can answer the questions about either insertion sequence or ISEScan.

@ Wirojne Kanoksilapatham, thanks for the comprehensive response. I have located my insertion sequence but i cant seem to locate the repeats that flank it. Is there a tool you can recommend I use or could you walk me through how to do that bit of my task?

Dear Kyanya,

I usually do the task manually. Based on the fact that the left and right inverted repeats (LIR and RIR) are usually located few bases away from the tnp gene sequences.

Approx 100 to 300 bp upstream and downstream should be included in your sequences. Then take the whole sequences (including the extra sequences) to blast again to the complete genomes (using highly similarity choice). After you get the results shown on highly similar sequences.

Locate all the possible sequences ( assigned the coordinate numbers) that matched to the query.

You may end up with 2 to 3 subgroups including the complementary sequences (+ and - strands).

From these sequences you may use any multiple alignment program to align the sequences. You may be able to see some direct repeat sequences (flanking the IS elements). Usually 3 to 10 bp-long. The IR sequences are adjacent to direct repeat sequences. This is a crucial evidence to locate some of the IS which you may be able to extend the finding.

Noted that not all the ISs contain direct repeats. The IR sequences are not always perfect match.

Also noted that common structure of an IS are shown as followed

direct repeat-LIR- promoter like sequence (approx 100 bp long with rbs upstream the initiation codon)- Tnp sequence (stop codon is often located inside the RIR)- RIR-direct repeat.

Dear Kyanya

The attached file is an example.

Hope it help you.

Wirojne Kanoksilapatham

Dear Kyanya,

Once you can locate the termini, you may fold the DNA or RNA.

If the sequence is too long. You should create an hybrid sequences between 5' and 3' (approx 100 bp 5'terminal sequence and 100 bp 3' end seq.).

Fold the hybrid. You may get several structure. But look for the one that show a stretch of complementary between both end.

The attached file contain a complementary sequences of both ends of an hybrid sequence of an unknown IS.

From the folding results (structure 11) (03-06-08-25_11) the LIR and RIR is proposed.

Putative LIR

5' GACCCTCCTCGTATAATAAGTTCGGTAAAAATTAAAATT 3'

Putative RIR

5' TATTTTTCTATTTTTTCTCCTCTTGACTTATTACCGGA AATC 3'

Cecilia Kyanya , the above response to your question contain an example to identify an inverted repeat sequence. Concisely, (1) identify left and right termini, (2) fold the left-end and right-end sequence using either DNA fold or RNA fold, (3) the IR is likely the partial duplex formed between the 5'- and 3'-ends.

Noted that direct repeat flanked the element is a crucial hint of the termini.

Wirojne Kanoksilapatham , Thank you so much for his elaborate description! It is very helpful!