Genboree

HELP TOPIC: "5. The LFF Annotation Format"




Show expanded help info?
 
5.1. Overview:

The Genboree LFF format is adapted from the LDAS upload format described at http://www.biodas.org/, specifically from the [ Annotations ] section.

These points are generally important:

  • The LFF format is tabular; each row is a single annotation record.
  • The annotation record is tab-delimited into 10 required columns, with up to 5 additional optional columns.
  • Regular spaces are allowed in many columns, because tabs are different than spaces.
  • NOTE: Do not use '{' or '}' characters. Due to a bug in MySQL's Java library code, in certain combinations, the data will not upload even though the data otherwise appears fine. MySQL is aware of this bug.
Thus, it is extremely similar to an MS Excel spreadsheet exported to a text file.
HINT:
  • Avoid LFF files with multiple sections; an annotation file should contain only annotations.
  • Use comment lines—whose 1st non-whitespace character is "#"— for example, to list column headers:
#class name type subtype chrom start stop strand phase score qStart qStop attribute-comments sequence freeform-comments


 
5.2. Column Descriptions:

The Genboree LFF Format has:
  • Ten (10) required columns:
    • - class, name, type, subtype, chrom, start, stop, strand, phase, score
  • Five (5) optional columns:
    • - qStart, qStop, attribute-comments, sequence, freefrom-comments

A detailed description for each column follows. For a more compact view, you can hide the Genboree context images and text using the Show expanded help info? above.

LFF Annotation Columns:
Col. #1:
class
  • - Required. Short text.
  • - A general 'category' for the annotation's Track.
  • - e.g. "Gene Predictions", "Conservation", "Repeats", "Assembly".

  • - Used to categorize annotation tracks;
    for example, below the browser picture:
Col. #2:
name
  • - Required. Short text.
  • - A name for the annotation/annotation group.

  • - All annotations with the same name are considered grouped.
  • - There are group-aware drawing styles that can suitably display such Annotation Groups.
  • - The exons in the 1st track all have different names and are probably not being drawn as the user would prefer.
  • - The exons in the 2nd track, however, are named according to their respective gene transcripts and can be drawn sensibly.
  • - Conversely, if all annotations are given the same name, they will all be in the same group. Group-aware drawing styles may not appear as you wish, and performance may suffer.
Col. #3:
type
  • - Required. Very short text. E.g. name or acronymn.
  • - The type of annotation; a repetition or a sensible sub-category of the class is best.
  • - Actually, any text you like, as long as it doesn't contain the ':' character.

Col. #4:
subtype
  • - Required. Very short text. E.g. name or acronymn.
  • - A more specific sub-type for the annotation; something more specific than type is best.
  • - Actually, any text you like, as long as it doesn't contain the ':' character.

  • - Together, the type and the subtype comprise the Track Name.
  • - To form the Track Name, they are joined by a ":". For example: "BCM" + "Novel Gene" = "BCM:Novel Gene"
  • - Ideally, the combined length of type and subtype should be no longer than 18 characters.
Col. #5:
  • - Required. Very short text.
  • - Name of the entry point (e.g. the chromosome) the annotation is on.
  • - It must be one of the entry points defined for the database.

Col. #6:
start
  • - Required. Positive integer.
  • - Start of annotation on the entry point.
  • - Start values beyond the ends of the entry point are prohibited.
  • - Note: the first base of an entry point is 1 (not 0). The start coordinate is included in the annotation.

Col. #7:
stop
  • - Required. Positive integer.
  • - End of annotation on the entry point.
  • - Stop values beyond the ends of the entry point are prohibited.
  • - Note: the first base of an entry point is 1 (not 0). The stop coordinate is included in the annotation.

Col. #8:
strand
  • - Required. One of: '+' or '-'.
  • - The orientation of the annotation with respect to the entry point.
  • - Use '+' if you don't care about strand.

  • - The strand is always available by left-clicking the annotation.
  • - Some drawing styles are orientation aware:
Col. #9:
phase
  • - Required. One of: 0,1,2 or '.'   ('.' == n/a).
  • - Whether the annotation is "in-phase" or "out-of-phase" with respect to something, such as the reading frame, or the other mate-pair read, etc.

  • - Currently, one drawing style is phase-aware: Paired-End
  • - Along with strand, it uses phase to visually indicate the relative orientation of mapped mate pair ends (i.e. whether the ends are in-phase or out-of-phase) when represented as a single annotation:
    • → ←  strand: +, phase: 0
    • → →  strand: +, phase: 1
    • ← ←  strand: -, phase: 1
    • ← →  strand: -, phase: 0
  • - The Paired-End drawing style does this by representing + oriented ends with a green block and - oriented ends with a yellow block:
  • - Other representations are possible, given user demand.
Col. #10:
score
  • - Required. Real number.
  • - A score for the annotation.
  • - e.g. 340, 0.871, 1e-10, 0, 1.0, etc
  • - We recommend "1.0" when score doesn't matter.

  • - The score is always available by left-clicking the annotation.
  • - Some drawing styles use the score directly.
  • - The minimum/maximum is globally-derived so the y-axis scale is uniform, regardless of location/view.
Col. #11:
qStart
  • - Optional. Integer.
  • - Start of hit in the query. Or '.' for n/a.

Col. #12:
qStop
  • - Optional Integer.
  • - Stop of hit in the query. Or '.' for n/a.

Col. #13:
attribute
comments
  • - Optional. A series of attribute=value; pairs.
  • - The attribute names are up to you, as are the values.
  • - Attribute=value; format is:
    •   · attribute name (any text not '=')
    •   · then '='
    •   · then value (any text not ';')
    •   · then ';'
  • - The attribute cannot be longer than 255 characters.
  • - If the value is longer than 65535, it will be truncated.
  • - This column can contain multiple attribute=value; pairs.
  • - Pairs found in this column are specifically modelled as 'attributes' or 'properties' of your annotation.

  • - These attribute-value; pairs have additional advantages:
    1. self-documenting comments with a regular structure
    2. easy to extract data into custom Link URLs
    3. looks similar to other formats (i.e. GFF)
    4. users looking at an Annotation's Details can make use of an 'auto-wrap' feature that makes reading such comments user-friendly
  • - LFF attribute-comment example:
gi=123987456; extDB_ref=10987K5; percIdent=94.68; e-val=1e-68; region=transmembrane; source=Smith Lab;
  • - Comment wrapping example:
Col. #14:
sequence
  • - Optional. Long text.
  • - This is intended for the sequence of the query or protein mapped to this region of the genome.
  • - Sometimes the query sequence and the genomic sequence are different (e.g. blating drosophila genes against sea urchin genome) and you want a place to put the query sequence.
  • - Be reasonable, however; not appropriate for storing the Mouse genome.
  • - Like comments, the sequence associated with an annotation will be available in the browser via left-clicking and choosing Annotation Details.

Col. #15:
freeform
comments
  • - Optional. Long text.
  • - We strongly recommend using the attribute comments to formally record extra content. It can be used for sub-selection, custom track links, etc.
  • - As a last resort, this free-form text column is provided.
  • - Be reasonable, however; not appropriate for storing War and Peace.


5.3. LFF Annotation Examples:

Minimal (10 columns; a 2-exon gene):

Genes & RNA	AVPR1A	Gene	RefSeq	chr12	63256962	63258172	-	.	0	.	.
Genes & RNA	AVPR1A	Gene	RefSeq	chr12	63260393	63263337	-	.	0	.	.

Standard (12 columns; contigs within a scaffold):

Assembly	AAGJ01021111	Assembly	Contig	Scaffold_114754	1	1300	+	.	1.0	1	1300
Assembly	AAGJ01022222	Assembly	Contig	Scaffold_114754	2195	3504	-	.	1.0	1	1310

With comments (13 columns; SNPs):

Cancer SNPs	HUR6.188	SNPs	Codon	chr2	19461847	19461847	+	.	0	.	.	allele=G/T; aaChange=A->A; nonSynon=false; refAA=A; mutAA=A; refCodon=GCG; rs_acc=rs123456; leftFlank=TGACGG; rightFlank=GCCAAC; exonPosition=2; proteinPosition=42; ampliconId=25299;
Cancer SNPs	HUR6.329	SNPs	Codon	chr12	19461988	19461988	+	.	0	.	.	allele=C/T; aaChange=Y->Y; nonSynon=false; refAA=Y; mutAA=Y; refCodon=TAC; rs_acc=rs987654; leftFlank=ACGCGC; rightFlank=GGGCGC; exonPosition=2; proteinPosition=89; ampliconId=25299;
Cancer SNPs	HUR1D.382	SNPs	Codon	chr18	22989108	22989108	-	.	0	.	.	allele=T/C; aaChange=N->N; nonSynon=false; refAA=N; mutAA=N; refCodon=AAT; rs_acc=rs789123; leftFlank=CTGTGT; rightFlank=GAAGAG; exonPosition=2; proteinPosition=360; ampliconId=25053;
Cancer SNPs	GRAF3.220	SNPs	Codon	chr19	36753139	36753139	-	.	1	.	.	allele=T/G; aaChange=S->A; nonSynon=true; refAA=S; mutAA=A; refCodon=TCC; rs_acc=rs567891; leftFlank=AGCTCC; rightFlank=CCGAGT; exonPosition=7; proteinPosition=310; ampliconId=24229;


5.4. Special Attribute=Value; Pairs:

Genboree will recognize certain special attributes within the attribute comments column (13th column).

Some of these are experimental, but the generally available and stable ones are listed below.

annotationColor
  • - Use this to set a color specific to this annotation.
  • - The annotation-specific color will override any track color settings.
  • - The color may be specified in the annotationColor attribute's value in one of 3 ways:
    1. RGB Hex Format: annotationColor=#FF00AA;
    2. RGB Dec Format: annotationColor=255,10,128;
    3. HTML Color Name: annotationColor=DarkGoldenRod;
      (see full list of color names)


 

 

 


Bioinformatics Research Laboratory
Genboree is built & maintained by the Bioinformatics Research Laboratory
at Baylor College of Medicine.
Genboree is a hosted service. Code is available free for academic use.
 
© 2001-2014 Bioinformatics Research Laboratory
    (400D Jewish Wing, MS:BCM225, 1 Baylor Plaza, Houston, TX 77030)
Questions or comments?