Microsatellite flanking regions are not necessarily unique sequences, but they may group into sequence families. This phenomenon seems to be widespread in Lepidoptera and also occurs in many other insect species (Meglécz et al. 2004). These microsatellites are likely to give multiple banding patterns during PCR amplifications, which can be very difficult to interpret. Therefore, identifying sequences that cluster together prior to primer design can save considerable time and money.
MicroFamily is a program designed for identifying flanking region similarities between different microsatellite sequences obtained from screening partial genomic libraries.
As a preparation for sequence comparison, sequences are edited by (i) replacing all characters other than ACGT by N (ii) by deleting the extremities if they contained more than two Ns in the ten most extreme base pairs, and (iii) by removing vector contamination if the sequence produced a BLAST hit against the UniVec vector base of NCBI (ftp://ftp.ncbi.nih.gov/pub/UniVec/). This latter step was not designed for precise removal of all vectors, but to avoid artificial similarities caused by vector contamination. The E-value (parameter that describes the number of hits one can "expect" to see just by chance when searching a database of a particular size) for the BLAST against the vector base can be specified by the user.
Microsatellites with motifs of 1-6 bp are identified. The minimal number of repeats for single base pair motifs and for all other motifs can be defined by the user.

The flanking regions of clean microsatellite-containing sequences are then compared by an all-against-all BLASTn analysis. The E-value can be specified by the user. Sequences are sorted into four categories based on the results of the BLASTn. They are classified as (i) Unique if no similarities were observed to any other sequences of the same dataset, (ii) Redundant, if the identity to another sequence was higher than 95% along the whole flanking sequence or (iii) UnBLASTable if the sequence had no hits at all (i.e. no significant similarity to any sequences), not even with itself. This is the case if the flanking region is too short or semi-repetitive (resembles to a microsatellite but there are not enough uninterrupted repeats). All non-redundant sequences that produced a significant hit with at least one different sequence were classified as (iv) Grouped.

Perl (ActivePerl 5.8.7.815)  

BLASTN-2.2.13   

CLUSTALw1.83 

Start by downloading and installing the above mentioned three programs. If you have a slightly older or newer version already on you computer, it will probably not cause any problem. Then download and unzip the appropriate MicroFamily zipped folder. Detailed instructions on the use of the program are found in the README file.

MicroFamily (822 ko)

Readme

MicroFamily

Readme (59 ko)

Former version (16 July 2006)

MicroFamily (820 ko)

Readme (63 ko)

Former version (16 July 2006)

MicroFamily (796 ko)

Readme (58 ko)