English
Français

Info

Help / FAQ

Summary

Activate the contextual assistance
The user interface

F.A.Q. (Frequently Asked Questions)

Main available tools

Questions about the tools

Questions about the data

Questions about the results

Questions about display

Other questions



Contextual assistance

To facilitate and accelerate the learning of this site, the user can activate the contextual assistance by clicking on this icon Activer l'aide located at the bottom left in the footer of the site. When the assistance is activated, the icon of origin is replaced by a new icon Désactiver l'aide which indicates that the contextual assistance is activated. By clicking again on this icon, the contextual assistance can be deactivated.

The contextual assistance corresponds to the display of tooltips when the cursor of the mouse passes to the top of certain zones like [this one]...and this is the corresponding tooltip for which information is available. Thus the displayed information corresponds to a context, i.e. that it refers to the zone where the cursor of the mouse is located. The tooltips thus make it possible to quickly learn how to use the site in situation without having to read beforehand all the documentation even if that remains always recommended.

↑ Go back to summary


The user interface

The RNA StrAT site is composed of 3 public sections:

The different colors are there to indicate to the user in which section he is and to help him remember in which section was the information he had consulted.


Information

This section contains documentation and various other types of information useful for the user. Here you will find explanations relating to the types of data used by this site and about how the algorithms worked. Certain sources of known skew are also explained there. Lastly, information concerning the site, who to contact to bring back a problem, to suggest ideas or to request information can be found in the sub-section "About".


Database

At this time, the database is only composed of the imported structures of RFam1. All the details concerning their importation are described in the sub-section "Algorithm Details". Included in this importation process are the multiple alignments, references to be quoted for each family of structures and also links towards the corresponding sequences in GenBank.

The structures of the database can be searched in various ways (according to the tree of families of RNA, according to the RNA families sorted alphabetically, according to the tree of species, according to the species sorted alphabetically or by RNAStrAT identifiers). The user can also search structures according to various structural criteria. When a list of structures is displayed, the user can select the structures which he wishes to compare with another structure and a simple on the comparison button (at the foot of the page) enables him to swich towards the tool of comparison (change of section). Moreover, each listed structure generally includes several links:

The page offering the details of a structure includes these links and others. It contains for example a link towards a multiple alignment containing the structure of interest or the links towards corresponding publications. It is also possible to visualize the structure in its "Fasta point-brackets" format by clicking on the button located below the structure.
Note: the RNAStrAT identifiers are, in theory, preserved by the updating process and can be used like permanent "accession" when linking towards this site.


Tools

This section contains a large variety of tools. There are tools intended for structural comparison, tools for displaying structures and various other tools. The use of certain tools is sometimes limited to avoid any voluntary or non-voluntary overload of the server by a user. Thus are limited: the number of structures which can be used simultaneously at the time of a request, the complexity generated by certain structure (number of free bases located on both sides of the same loop) and the numbers of requests that can be carried out simultaneously by the same user (identified by its IP address). If for some reasons, these limitations are a problem for you, you can contact us and we will possibly be able to modify those limitations according to your needs, if they are justified.
Note: the services offered by this server are free and anonymous and the limitations are there so that everyone can benefit from it in an equitable way and under good conditions.

The program used for structural comparison is the "RNA Structure Alignment Program" (RNAStrAP). This program was developed by our care in C++ and the Windows and Linux versions (binary and sources) are available for download on demand.
The display of structures (alone or aligned) is carried out by two tools: the first is RNAPlot2, which is only used to calculate the position of the bases on the image; the second is the engine "RNA Strat Secondary Structure Rendering Engine" (RSSSR Engine) which we produced in php and which is used to colour the bases, to add various graphic elements and to display the result under various formats.
Note: In the next release of the website, the comparison and display tools will accept requests coming from other servers.


Technical notes

This site is lodged on one of the LBIT server. Technologies used are:

This site was tested and is displayed normally with the navigators Firefox 2, Internet Explorer 7, Opera 9, Safari 2 and Konqueror 3. It conforms to the standards XHTML 1.0 Transitional, CSS2 and Javascript 1.4 and higher. If your navigator does not display correctly the pages of this site, that can be because it does not supports one of these standards or that it is badly installed or configured or that our site presents some disfunctionnality. If you are convinced that it is indeed a disfunctionnality of this site, you can bring back your problem to us by specifying its nature, the way of reproducing it and your configuration (OS and navigator with their respective versions). We will then try to resolve the problem.
Note: page caching on some browsers can sometimes raise some issue and if you face something abnormal, it is recommended to disable caching and try again.
Note: the site can function when Javascript is not available; the user will simply have to carry out certain operations manually (like the refreshing of the pages) and the results will be presented in a less friendly way (no sorting or presentations in tables). On the other hand, if the Javascript functions work only partially, the quality of display of the results will be random and in certain cases, the user will not be able to consult them. So, in these extreme cases, it is recommended to completely desactivate Javascript.

↑ Go back to summary


Pairwise comparison

This tool enables to compute the edit distances and an alignments between an RNA secondary structure provided by the user and a set of at least one RNA secondary structure. This set of structures can be either provided by the user in a file to upload or selected from the database of available RNA secondary structures. Algorithms details can be found in the section Algorithm Details.

↑ Go back to summary


Distance table

Given a user-provided set of RNA secondary structures, this option computes a table of pairwise edit distances/similarity scores.

↑ Go back to summary


Database search

This tool allows the user to search in the database for the set of RNA secondary structures that are the closest from a user-provided RNA secondary structure. There are two options: searching for complete structures of the database or searching for substructures (stems and stem-loops) of structures of the database.

↑ Go back to summary


Rendering

This tool allows the user to display, navigate and export pictures of RNA secondary structures and RNA secondary structures alignments.

↑ Go back to summary


What are the differences between an edit distance or a similarity score for an alignment?

An edit distance is a value which reflects the minimal total cost of the operations necessary to transform a structure into another. This value must always be positive or null. This implies that the costs associated to the different edit operations have to aso be positives or null. The weight associated with each edit operation corresponds to the evolutionary cost which one allots to such an operation; a null cost indicates that the operation does not have any impact while a high cost indicates that this operation is not very probable. The bigger is the distance between 2 structures, the bigger is the divergence between these structures and conversely, if their distance is small, then they are close. A distance of 0 between 2 structures indicates that they are the same structure.

The calculation of an alignment's score turns over a value that indicates to which point two structures are similar. Contrary to the edit distance, a score can be positive or negative. Thus, the weights associated with each edit operation can also be positive, null or negative following if one wants respectively to support, make neutral or disadvantage an operation. According to the score scheme used, i.e. the set of the different scores associated with each edit operation, the alignment's score can emphasize or not structures comprising a certain type of similarity (for example a higher score for the correspondences of pairs G-C than for the pairs A-U will emphasize, if both structures have an equal number of pairs, the structures having more pairs G-C than pairs A-U). Generally, the more the alignment's score is high, the more the structures are similar and conversely, the more it is low, the more they are different.

From a purely computational point of view, computing the edit distance or score of an alignment are one and the same, the only difference being that in the first case one seeks to minimize a value and in the other to maximize it. From a practical point of view, one will rather use, for example, the computation of distances when studying the divergences between several species and the computation of scores when the similarity of several structures is studied. In conclusion, these 2 types of computations carry out the same results but the use of one rather than the depends on practical aspects and the problem we are dealing with.

↑ Go back to summary


What is a score scheme? What is the meaning of the different edit operations?

The score scheme is the set of the different weights (costs) associated individually to each edit operation. An edit operation corresponds to an evolutionary event which would have made it possible to transform a subelement of a structure into another. The literature counts a certain number of such operations to which we added some which seems pertinent to us. Certain operations cannot be detected by current algorithms and it also should be known that according to the score scheme employed, certain operations can disappear because a combination of others, in that scheme, is more probable. For example, if we are computing an edit distance, the alteration of a base pair will never be detected if this operation is cost more than the removal of the pair followed by the insertion of a single base (free base).
Here is a list of all our "edit operations" with their associated symbol (found in the alignment scenario), an example and a brief description:

m
m match a free base
s
s substitution of a free base
i
i insertion of a free base
d
d deletion of a free base
M
M match of a base pair
H ou h
h half-match of a base pair
S
S substitution of a base pair (both bases in the pair are subtituted)
I
I insertion of a base pair
D
D deletion of a base pair
C ou c
C completion of a base pair: a new base is inserted in the structure and naturally forms a pair with another base already present
P ou p
P creation of a base pair (pairing): the formation of an hydrogen bond between to existing free bases
A ou a
A base pair altering: deletion of one of the two bases of a base pair resulting in the other base becoming free
B ou b
B breaking of a base pair: deletion of the hydrogen bond between the two bases of a pair resulting in two free bases
R ou r
R rearranging one base of a base pair: the base "partner" changes
t
t insertion of a stem (tige) or stem-loop
T
T deletion of a stem (tige) or stem-loop
j
j connection (join) between two segments of a sequence: formation of a covalence link
l
l split of a sequence into two segments: breaking of a covalence link
F ou f
folding of free bases: the bases are free but have a tendancy to form base pairs
W ou w
change of side of a free base (side swap): a base in one orientation (for example 5' -> 3') will find itself in the opposite orientation (3' -> 5') on the opposite strand

The edit operations can be seen from a general point of view as in the list above but also from a point of view specific to the nature of the bases brought into play. We can for example consider that to break a pair A-U (weak hydrogen bond) costs less than to break a pair G-C (strong hydrogen bond) even if it is the same type of operation. On this site, we thus will distinguish the costs associated with the type of operations and the specific costs associated to the bases brought into play. When the user specifies particular costs depending on the bases, the cost of an operation will be calculated in the following way: if a cost was specified for this operation on these bases, it is used, if not the "by default" cost of this type of operation is used. For example, if the user specifies that the substitution of a pair by another costs 2 but that the substitution of A-U into G-C costs 1.5, any substitution of A-U in G-C will cost 1.5 but substitution of U-A in C-G or G-C in A-U (the order is important here!) will cost 2.

↑ Go back to summary


How can I align my structure with some structures from the database?

Three methods can be used to achieve this task. The first consists in traversing the database, selecting the structures, the RNA families or the species of interest (by selecting the boxes of the corresponding lines) and to use the button "Pairwise comparison" located in general in the bottom of the pages displaying the lists of structures. This will cause to go in the structure comparison page by including the selected structures in the list of structures to be compare. The following methods operate directly from the comparison page (or the one computing table distances).
The second method concist in the use of the link "+ Select structures from database" that open a popup window which allow you to traverse and select the structures to be compared (while clicking on the link "+ Add selected entries").
Finally, the last method concist in entering the RNAStrAT identifiers ("RSxxxxxx") of the structures to be compare (one identifier by line) into the text zone of the "Structure(s) for comparison".
Note: the number of structures which can be compared during a comparison is limited: when RNA families or species contain more structures than this limit, the one that exceed the limit are ignored.

↑ Go back to summary


I'm running some long calculations but would like to be able to close the window and return to the results later on. How can this be done?

Calculations are not dependent on a session but on a calculation identifier which is provided in the URL (addresses) results's page (the one that is refreshed automatically as long as the calculations are not done). You can thus simply back up the address of the page (bookmark) to be able to return there later. You should only verify that this address contains one follow-up idenfier of the form "tracker=<random list>" (e.g. "tracker=cfed5f420812efd9a514f6c614d8d1fekanicfat"). However, you have to be aware that your results of calculations are not preserved indefinitely (the minimal time of conservation is indicated at the bottom of the comparison page, it is 168 hours; passed this delay, a result may be deleted).

↑ Go back to summary


What are the computational limitations?

On this site, several computational limitations are used for various reasons. First of all, a user (identified by his address IP) cannot simultaneously carry out more than 3 requests. In each request the number of structures to compare cannot be more than 1000. In the distance tables computation the number of structures is limited to 200. To avoid computations that are too long or lack of memory problems, the complexity generated by the comparison of two structures is calculated in the following way: complexity=A*B*C*D, where "A" is a number of free bases on one side of the larger loop of the first structure, "B" is the number of free bases on the other side of this loop, "C" is a number of free bases on one side of the larger loop of the second structure and "D" is the number of free bases on the other side of this loop (in the case of the terminal loops, we consider that each side contains the total number of free bases of the loop). The generated complexity cannot exceed 12960000 and if so, the comparison of these structures is cancelled. The sizes (in bytes) of the images are also limited (but this limit should not be reached in theory without alterating the forms). Finally, the minimal storage time of the results is 168 hours. Passed this delay, it is possible that out-of-date results will be removed.
Note: some of these limits can be modified. If the imposed limits are not appropriate to you, you can contact us by justifying your needs (and by providing your IP address) and we will carry out, if possible, the needed modifications.

↑ Go back to summary


How do database search tools work?

Database search algorithms details are provided on the page "Algorithm Details".

↑ Go back to summary


How are RNA secondary structures define on this site?

To what correspond the different elements of a structure?

What is a stem or a stem-loop?

You will find information answering these questions in the Data Specification section.

↑ Go back to summary


What is the IUPAC/IUB code?

The IUPAC/IUB code is a standard used to represent by letters one or more bases. You can find details on this standard at this address (PDF) or this one (HTML). 16 diffenrent codes are defined: A, B, C, D, G, H, K, M, N, R, S, T, U, V, W et Y.
A, C, G et U, represent respectively the bases Adenine, Cytosine, Guanine and Uracil. All the other codes represent more than one bases. For example, R reprensent the purines A and G. When such codes are generated, the comparison tools considered them as being symetric, i.e for example it consider A to match R and R to match A. This symmetry can be broken by modifying the score scheme of the considered bases. The recognized correspondences between the various codes are given in the following table ("+" for a match and "-" for a substitution):

 GARCSMVT/UKWDYBHN
G+-+-+-+-+-+-+-+
A-++--++--++--++
R+++---+---+---+
C---++++----++++
S+--++-+-----+-+
M-+-+-++------++
V+++++++-------+
T/U-------++++++++
K+------++-+-+-+
W-+-----+-++--++
D+++----++++---+
Y---+---+---++++
B+--++--++--++-+
H-+-+-+-+-+-+-++
N+++++++++++++++

Note: on this site, the base T (thymine) is automatically converted to U (uracil).

↑ Go back to summary


Is it possible to use other symbols for bases then A, T/U, G or C?

It is only possible to use the codes in the l'IUPAC/IUB standard but these codes have some very specific significations. If you want to use, in your structure, some codes that corresponds to non-standard bases (inosine, for example) it is possible to replace those bases by some IUPAC/IUB codes that you do not use (if you don't already use all the 16 codes). You will then have to redefined all the operations in the score scheme that uses this non-standard base to avoid matches that are not wanted. The alignement process can be done just like if there was an additional code but be aware the alignment scenario annotation may not reflect what the alignment shows. For instance, let's say you used the "Y" (pyrimidines) code for inosine and in your alignment a "Y" is aligned with a "C" (cytosine), these aligned bases will be annotated by a match ("m") even if it is in fact a substitution. This artefact is due to the annotation method that finds a cytosine matching a pyrimidine (detection of match/substitutions is not based on the score scheme) but the score/distance computation will reflect the substitution (according to your score scheme).

↑ Go back to summary


What is the reliability of the structures in the database?

You will find the answer to this question in the section "Algorithm Details" (sub-section "Structures importation").

↑ Go back to summary


Why is one of your tool doesn't want to accept my structures?

There are different possible reasons and it is your responsability to find which one. The first reason is that the data your using is in an incorrect format. The different formats recognized by this ite and their specifications are described in the section "Data Specification". Another possible problem is that you gave to many structures to the comparison tool. There is a maximal number of structures by request of comparison. If there are too many structures the ones that exceed the limit will not be rejected, they will simply be ignored. It is also possible that the structure you want to compare with others have a particular form that implies to much calculations (cf. "What are the computational limitations?") and this structure will by ignored by the comparison algorithm. Finally, all structures with pseudoknots cannot be compare with our tools.

↑ Go back to summary


How can I interpret the scores and alignments?

Two types of calculations are possible: the computation of an edit distance or the computation of a score. If your computing a distance between two structures the more its value is high, the more the structures diverged. If the distance is null, this means that the structures are equivalent (or that there is a computational problem!) . Il you are computing an alignment score between two structures, the more its value is high, the more the structure are close or similar and, conversely, the more its value is low, the more the structures are different. The alignment that is returned by the different tools is one of the optimal alignment between the structures i.e that, for a given score scheme, the tools return the alignment that have the minimal distance or maximal score. This alignment indicates which element of a structure corresponds to which other element on the other structure. An alignment scenario is also normally given and this scenario explain which transformations is needed to transform the first structure into the other.
Note: comparing alignments scores of structure datasets does not really make sens because scores are intimately linked to structure dimensions (i.e. with a same mutation rate, "big structures" datasets will have higher distance than "small structures" datasets). For instance if we compare the distance of 5S rRNA between taxa A and B and the one between SSU rRNA 5 of the same taxa, distance between 5S rRNA is expected to be smaller thant the one between SSU rRNA 5 especially is the 2 kinds of structures evolved at the same speed: the distance (or the score) is not an evolution rate indicator.

↑ Go back to summary


Up to what point are the calculations of scores or distances reliable?

The reliable of the scores or distances is equivalent since the algorithms used in both cases are practically identical. At this moment, no statistics are provided to make it possible to evaluate the reliability of the provided scores. Despite this fact, one can consider that the calculated scores are completely reliable since they are exact (we are using algorithms that compute exact scores or distances and not heuristics). It is however necessary to pay attention to the selected score scheme (the weights associated with the operations must have a biological or logical sens) and to the fact that the calculated score corresponds to the score of a global alignment of the structures and not only to one of a simple stem/stem-loops correspondance.
Note: other algorithms are now being implemented to compute a stem/stem-loops correspondance score when a global alignment of the structures is not possible.
Note: we are now studying the possibility of introducing e-value and p-value for secondary structure alignments.

↑ Go back to summary


Why is it that in certain cases I can not obtain global alignments but only stem or stem-loops correspendences?

This is a problem related to the way our algorithm works. The algorithm decompose the secondary structures into stem/stem-loops and compare them between themselves in order to find the best stem/stem-loops correspondance. The order in which the stem/stem-loops are compared comes from a traversal of the structures from the 5' to the 3' extremities. By this method, information on the layout between the stem/stem-loops of a structure is lost and when the structures are very different, that can lead to the impossibility of finding a global alignment. One way to correct the problem will be to do an hierarchical traversal of the stem/stem-loops instead of a linear one so that we could keep track of the layout of the stem/stem-loops (to do that we could put the stem/stem-loops of the two compared structures into trees and then compare these trees between themselves). Another more radical solution would be to completely change our algorithm. Some work is now being done in these two different directions.
Note: This problem usely happen only if we are trying to align really different structures.

↑ Go back to summary


I'm able to launch my calculations but the display of the results is strange. What can be the problem?

If your are not using one of the tested browsers (FireFox 2, Internet Explorer 7, Opera 9, Safari 2 et Konqueror 3), it is possible that the one you are using doesn't correctly implement the different standards that we use, causing some display problems. However, if you are using one of the tested browser there can be different causes for your problem. That can be due to a calculation error, to out-of-date scripts in the mask or other problems related to the mask. You can try to manually refresh the page and if after that nothing have change you can bring back the problem to us by specifying your configuration (OS and browser with their respective versions) and if possible the link towards the results page.

↑ Go back to summary


When I compare two structures and I change the order of comparison, the score obtain is the same but the alignment is different. Is it normal?

Yes! In fact, this case happen when there is more then one optimal alignment between two structures. Changing the order of comparison of the structure may change which optimal alignment is returned.
Note: in futur developpement we are envisaging to return the set of all optimal alignments between two structures. In that case, both order of comparison will have the same output.

↑ Go back to summary


I would like to use some of the results or data given by this site. How do I proceed?

The result's pages of the various tools generally propose a way to export the data in a text format, which should facilitate your live. For the comparison's results, the results of a search, or the computation of distance table, the results can be selected and copied in a any spreadsheet tool, like Microsoft Excel and Open Office Calc. For these two particular tools, it is recommended to use the "Paste special..." function and to use the "unformated text" format so that the column breaks are recognized automatically. For stems/stem-loops search results, it is possible to get the line breaks back into the alignments by using an appropriate tool that can replace the character of hexadecimal code 11 by a line break (code 10, 13+10 or 10 depending the operating system). In all the cases, it is recommended that you also include in your possible publications the score scheme employed for your calculations. Once the results are correctly formatted by your care, do not forget to quote this site but also the publications which provided the structures that you used. These publications are generally related to each family of structures and you will find a reference to them on any page describing one member of the structure family.

↑ Go back to summary


How can I save my computations to be able to look at them again later on?

To back up the results of your submitted calculations, you can, for the moment, only use the links permitting you to export the various types of results into text format. However, if you try to back up the popup window containing the results you will have the impression that everything is nice and saved but in fact nothing will be saved! Indeed, the contents of the popup windows are generated by a script and does not correspond to any real web page. As the recording process of the majority of the browsers seeks the page of origin when one tries to back up the page in progress, it will not find anything since this page does not have a URL (an address) and consequently will back up an empty document or the contents of the page parent of the popup! To avoid this problem, you must rather select all the results and copied it into your favorite word processing tool. You will then be able to save your data.
Note: Why is it that complicated? It is because, for some computer security reasons, Javascripts scripts are not allowed to save files (if it was not the case, this method would have been employed).
Note: you can also back up the raw output of the algorithm but for the moment, the site does not make it possible to treat the raw output to present it into a table.

↑ Go back to summary


It is possible to see how the distance or score is compute between two structures?

Yes, by checking the box "algorithmic details in raw output" in the comparison page and by downloading the raw output on the results window. This output will contain the distance tables and various information on the backtracking of the alignment. Its reading should remains, however, an experts fun!

↑ Go back to summary


How can I obtain the other potentiel optimal or sub-optimal alignements?

For the moment, it is not possible. However, it is expected that in a forthcoming version, other optimal alignments will be provided.

↑ Go back to summary


Can I built some multiple alignments from some alignments of pair of structures that I have already obtained?

At this time, no tool on this site is able to do this. You will thus have to turn towards other software or to carry out this task by hand... or wait until we develop this kind of tools, which can take a certain time.

↑ Go back to summary


Your site does not seems to be correctly displayed on my computer, what can I do?

If your are not using one of the tested browsers (FireFox 2, Internet Explorer 7, Opera 9, Safari 2 et Konqueror 3), it is possible that the one you are using doesn't correctly implement the different standards that we use, causing some display problems. If some Javascript functionality seems to be working well both others don't you can try to completely desactivate Javascript to see if the diplay becomes better. However, if your browser is part of the tested ones, you can report your problem by describing it in details (possibly with a screen capture to help us understand what is happening) and by specifying your configuration (OS and browser with their respective versions).

↑ Go back to summary


What can I do when the display of a structure does not seems to answer to the commands of the navigation panel?

Try to desactivate the functions of the mask of your browser. If you do not find the option allowing you to desactivate them, try to empty it. If you don't know how to do that, close your browser, open it again and go back where you were. Unfortunately, I did not find yet a miracle solution to this problem but I'm working on it.

↑ Go back to summary


What are the different types of alignment rendering?

How can I interpret these rendering?

There is the rendering of an alignment of two structures and the rendering of consensus alignments. There are 6 types of alignment rendering between two structures:

Types 1 and 2 render the first or second structure that you want to align annotated with all the edit operations that will change this structure into the other one if the alignement. A circle surrounding a base inticated a substitution and the color of the circle inticates the nature of this substitution: blue indicates a substitution by an adenine, yellow by an uracil, green by a guanine, red by a cytosine and black by an undetermined base (cf. IUPAC/IUB codes). Crossing a base means that it is deleted. One or several bases in a green box inticates an insertion site. Finally, if an hydrogen bond is red, then it has to be broken and if it is orange, it indicates a base altering. The creation of an hydrogen bond is not represented on this type of display.

Types 3 and 4 are different from the first two ones by the fact that all the bases in an insertion sites are included in each structure and not added over like annotations. Thus, that makes it possible, on one hand, to see all the inserted bases framed in green and on the other hand to display in green the created hydrogen bond and in turquoise the bonds formed by the completion of a pair. The removed bonds remain indicated in red for suppression and orange for alteration. The removed bases are framed in red and the substituted bases are in blue frames, the first being the base of origin and the second the one resulting from the substitution.

Type 5, "maximal commun sub-structure" display the structure who contains only the bases and bonds that are common to both structures. For the moment, the substituted bases are framed in blue (the 1st base is that of the 1st structure and the 2nd, that of the 2nd structure) but it is envisaged to replace them by their corresponding IUPAC/IUB codes.

Type 6, "minimal super-structure" is, at this moment, equivalent to type 3 but like it was the case for the type 5 the substituted bases will be soon replaced by the corresponding IUPAC/IUB codes and another type of annotation will be used for the transferred hydrogen bonds.

A consensus alignment results from aligning one structure independently to several others, i.e. it is not a question of a multiple alignment but of multiple alignments. This consensus alignment makes it possible, for example, to emphasize on a structure of interest its differences or commun points with a structure family. The annotation of the structure is to some extent a statistical representation of alignments. Each circle around a base indicates that this one was aligned with other bases of the other structures and the proportion of each type is given by the size of each arc. Thus, an arc of 120° of a certain color indicates that the base encircled was aligned in 1/3 of the cases (120° = 360°/3) with a base whose type corresponds to the color of the arc (blue=adenine, yellow=uracil, green=guanine, red=cytosine and black=code of different types of bases IUPAC/IUB). The deletion rates of a base is indicated by the intensity of a red cross on this one. The more intense the red is, the more the base was often removed (i.e aligned with a hole). A base which is not encircled is a base which is never aligned with another base on the other structures, therefore this one is removed in 100% of the cases. Lastly, the last statistical annotation is that of insertions: when there was at least an insertion between 2 bases or at the ends of the structure, a number is displayed at this place and corresponds to the average number of bases inserted compared to the totality of alignments. Thus, for an alignment of a structure with 4 others, a "0.5" on the covalence link between 2 bases indicates that there are on average 1/2 base inserted into this place, which can be interpreted in several ways: 1 base was inserted for 2 structures and none for the 2 others or 2 bases was inserted on 1 structure and none on the others. For the moment, no statistical annotation is displayed concerning the hydrogen bonds but we are thinking of doing it.

↑ Go back to summary


I have some comments about your site but I'm afraid to misuse your time. Should I ask anyway?

Of course! If this question is still there, it means that for the moment no significant abuse occurred. Our goal is to offer one of the best sites of secondary structures comparisons. You use this site and if your comments can help with its improvement everyone will benefit from it, including yourself, so do not hesitate any more. All the comments are good, even the most naive ones.

↑ Go back to summary


What is...the Universe?

How come is there a fun question here? A little relaxation and phylosophie is not bad after all!? Initially, the answer to this question can only be personal and subjective because if we considered that the Universe contains "all" and that nobody holds infused science, certain elements will obligatorily escape us and make our analysis of the Universe so incomplete that we will not be able to define it. In this context, I can now present my point of view on the question knowing that you have the right to have a different opinion. The Universe exist (with the "common" meaning of the term in opposition to a "conceptual" meaning by which all and anything could exist) only because we are there to observe it. The Universe thus depends on our existence. So the question now is: "What is our existence?" Two aspects emerge from this question, the origin of our existence and towards what it is moving. Since the future is dubious, I will not try to answer the second aspect of the question. So I will fold back on the first: where do we come from? Several theories exist and the most credible and interesting one, in my opinion, is based on the theory of evolution. According to this theory, we all comes from the LUCA cell. But how did this cell appeared? Some thinks that form the initial "primitive soup" of the Earth, organic molecules could have been formed by chance (under the action of the environment and certain chemical conditions) and it is proposed that some of them were able to auto-replicate and to gradually produce other organic molecules with which an association was possible, finally leading to a cell embryo. We can now ask ourselves the following question: "To what did these molecules looks like?" And I believe in what many well known scientists suggested: to molecules close to the RNA molecules we have today. Indeed, certain RNA molecules are able to polymerize other RNA molecules, the RNA molecules can be used as genetic data carrier and finally, many RNA molecules are able to catalyse a large variety of chemical reactions.
In conclusion, without the RNA the Univers would not be a big thing and this site is there to answer the question which you must ask youself now: "What is that RNA?"

↑ Go back to summary


I saw the RSS link, it seems cool but I have no idea of what it is? For what it is used?

The RSS stream of this site contains all the advertisements of the site like its innovations, problem solving or other information. Registering to the RSS stream is the best way of being held informed of the evolution of this site. The operation is similar with that of a mailing-list in the sens where the news on the site is received as mails, if your mail provider supports RSS streams. However, RSS streams have some huge advantages against mailing-list: first of all, your email adress is never used nor given (it is your mail client that will connect to the site to see if there is some news). This guaranteed you no spam problem related to your inscription with the RSS stream. That also facilitates your inscription since no data concerning you is required and finally you can at any time desactivate your inscription and this desactivation dosen't have to be validated by an administrator.

↑ Go back to summary


My question does not seems to be on that list, what should I do?

Your question doesn't appear here? If you think that it is nevertheless relevant, you can write to us by using the mail address provided (in the footer on the right).

↑ Go back to summary

  1. Griffiths-Jones, S., Bateman, A., Marshall, M., Khanna, A., Eddy, S.R., RFam: an RNA family database, Nucleic Acids Research, 2003, 31, 1, 439-441.
  2. Ivo L. Hofacker, Vienna RNA Package.