• Description
• Organisms and Libraries
• Technical Notes
  • Program output
  • Genome coverage of a contig
  • Genome assembly affect
  • Segmental duplications in human genome
• FAQ

Description

BAC contig assembler is a computational tool for assembling contig of minimally overlapping BAC clones across genomic sequence of interest for mouse and human genomes. Resulting minimal tiling path BAC contig covers up to 99% of region with average 1.3x  genomic representation  and BAC density of 7-8 BAC clones per 1MB of sequence.

BAC contig assembly is based on WU-BLAST search of genomic sequence against TIGR BAC End Sequences Database . Availability of both BAC end sequences allow one to map clone onto genomic sequence with the help of BLAST algorithm. Using TIGR database results in retrieving redundant set of BAC clones mapped to the sequence (Fig.1A). BAC contig assembler algorithm uses this redundant set of clones to assemble minimal tiling path BAC contig which is comprised of non redundant subset of minimally overlapping clones (Fig.1B).

Fig.1 BAC contig assembly strategy.
A) Mapping all BAC clones from TIGR database onto genomic region of interest. B) Minimal tiling path BAC contig assembled with BAC_assembler.pl Perl program.
BAC clones are mapped with WU-BLAST and represented by thin horizontal bars joining short thick bars which are the position of BAC end sequences in the query sequence.
Taking into account the average BAC vector insert size of 150-200kb  we use only those BAC clones which end sequences are mapped within 300kb from each other allowing for short gaps of uncertain size within genomic assembly.
X axis represents nucleotides of genomic sequence.
Visualization is implemented with Genome Cryptographer software.

Organisms and Libraries

BAC contig assembler program allows for construction of minimal tiling path contigs across regions of mouse and human genome assemblies generated by UCSC Genome Project .

Following BAC libraries may be used for contig construction:

• For human: CIT978SK (A B), CIT-HSP (C D1 D2), RPCI-11
• For mouse: RPCI-23, RPCI-24

Technical Notes

• Program Output  is a set of gif and text files. If sequence of interest is larger than 10MB, it's partitioned into parts of 10MB size which overlap in 500kb. Then a set of output files is generated for each part of a sequence. Text files contain list of BAC clones with start and end coordinates within the query sequence, chromosome number, start coordinate within the chromosome, range of chromosomal coordinates, pair of genomic coordinates  in whole genome scale. Contig characteristics is listed at the bottom of a text file.

  
  ==========================================================
  query name "name_of_the_sequence"
  UCSC assembly

  RPCI11_R-135P13 18295 179497 8 123718296 chr8:123718296-123879498 1506910454 1507071656
  RPCI11_R-580G9 130099 302656 8 123830100 chr8:123830100-124002657 1507022258 1507194815
  CIT-HSP-D_3218D9 288803 323966 8 123988804 chr8:123988804-124023967 1507180962 1507216125
  RPCI11_R-755E24 383461 564016 8 124083462 chr8:124083462-124264017 1507275620 1507456175
  RPCI11_R-37N22 563889 745693 8 124263890 chr8:124263890-124445694 1507456048 1507637852
  RPCI11_R-701E22 702226 893282 8 124402227 chr8:124402227-124593283 1507594385 1507785441
  .......
  1. the query length: 10000000 bp
  2. the number of BACs: 79
  3. mean value of BACs overlap.: 34.31818 %
  4. coverage: 95.90789 %
  5. genome representation: 1.3250137 x
  ==========================================================
   

• Genome coverage of a contig depends on  repeat content of query sequence, genome assembly quality (see below) and representation of a sequence in the library. High repeat density precludes from mapping BAC end sequences onto genomic sequence which may  lead to gaps in a contig. In addition some sequences are recalcitrant to cloning which results in their under representation in genomic libraries and hence may also lead to contig gaps.
• Genome assembly has an influence on contig quality. Contig is built up of clones which ends map within the 300kb from each other (for deatails see fig.1 legend). Clones which do not meet this requirements are excluded from assembly. Therefore gaps in a contig may stem from the wrong assembly of a region (for example see fig.2).

Fig. 2. Assembly affect on contig construction. Thick bars represent BAC end sequences mapped onto query sequence. In some cases the length of homology is far more than the one of BAC end sequence which might be due to presence of repeated stretches of unique DNA  in local region of genomic assembly. These clones are not considered for contig assembly.
 
 

• Genomic segmental duplications comprise about 5% of a human genome. Genome wide analyses using BAC contigs such as aCGH might be affected by segmental duplication content of genomic region which can be estimated with the help of Segmental Duplication Database ( J. Bailey et al, Science 2002 Aug 9;297(5583):1003-7.)

BAC contig assembler FAQ

• Is it possible to assemble library specific contig?
• How can I fill in gaps in a contig?
• No contig has been assembled across genomic sequence?

Is it possible to assemble library specific contig?

BAC contig assembler program allows one to construct both library specific and mixed contigs. See Section 3 on submission page.

How can I fill in gaps in a contig?

Absence of sequence in a library causes gaps in a library specific contig. In this case gaps might be filled in with clones from other libraries. Click Yes in Section 7 on submission page to have redundant set of BACs mapped to the sequence of interest. Use  "All BAC clones mapped to sequence_name" and "List of all BAC clones mapped to sequence_name" files to insert necessary clones into minimal tiling path contig.

No contig has been assembled across genomic sequence?

There might be several reasons for absence of any clones spanning the region of interest: (i) high repeat content of sequence; (ii) there's no sequence available in the range of coordinates (you may check your sequence with the help of UCSC Genome browser. Note, that coordinates are freeze specific! ); (iii) low quality of assembly in the genomic region; (iv) the region is recalcitrant to cloning which leads to its absence in genomic libraries.