castanyes blaves
Random ramblings about some random stuff, and things; but more stuff than things -- all in a mesmerizing and kaleidoscopic soapbox-like flow of words.5/11/2009
Anolis carolinensis: First reptile in Ensembl -- Amonida Sadissa
- Very few anolis proteins and cDNAs, many more ESTs.
- Used Uniprot PE Evidence ranking to generate transcript models with genewise
- Different parameters for Exonerate, including exhaustive option for cDNAs, including 31000 chicken set (which wasn't very useful in the end)
- Chris Ponting group provided extra models and kill list to rename some retrotransposons and pseudogenes
- Manually looked at some of the EST genes in Chris' list to reincorporate them in the main db
Labels: ensembl, gene prediction, scientific talk
4/30/2009
RNAseq in the worm -- Gary Williams
Illumina Short-read transcriptome data has the potential to help solve many problems with curating gene models and the genomic sequence in C. elegans. This is an initial look at the data and some examples of how it can be used.
So far, C. elegans gene predictions:
36% - fully confirmed by ESTs
48% - partially confirmed
14% - no transcript confirmation
RNAseq data -- different worms than the genome, so some polymorphisms expected -- 200bp inserts, 36bp paired end reads
MAQ or cross-match to genomic or transcript sequences
6137 new splice junctions (6% increase)
Jumped from 70000 to 98000 splice junctions.
3x as many polyA sites
80 possible new coding genes
V-shaped coverages -- validation against traces, then:
Updated gene builds will be given to other projects. Next Ensembl Metazoa comparative genomics build may already have the modENCODE-updated C. elegans and D. melanogaster builds.
So far, C. elegans gene predictions:
36% - fully confirmed by ESTs
48% - partially confirmed
14% - no transcript confirmation
RNAseq data -- different worms than the genome, so some polymorphisms expected -- 200bp inserts, 36bp paired end reads
MAQ or cross-match to genomic or transcript sequences
6137 new splice junctions (6% increase)
Jumped from 70000 to 98000 splice junctions.
3x as many polyA sites
80 possible new coding genes
V-shaped coverages -- validation against traces, then:
- Detected sequencing error, correction needed for the reference
- Detected alternative haplotype
Updated gene builds will be given to other projects. Next Ensembl Metazoa comparative genomics build may already have the modENCODE-updated C. elegans and D. melanogaster builds.
Labels: ensembl, genomics, nextgen sequencing, scientific talk
4/23/2009
Phenotype data in Ensembl -- Fiona Cunningham
European Genotype Archive: Genome Wide Association Studies (GWAS) like WTCCC data and others. Only public information is public available under very strict rules.
NHGRI GWAS will be imported in Ensembl: it's got manually curated data of high quality.
Links in Variation view: link "Phenotype Data (n)"
http://www.ensembl.org/Homo_sapiens/Variation/Phenotype?source=dbSNP;v=rs420259
Diagnostic testing: situation will now improve in Ensembl
Locus specific databases (LSDBs): p53, ABO, collagen, albinism, cystic
fibrosis, Altzheimer's disease, ... >700
The main aim is to be able to link the reference CDS sequence used by the biomedical community to the most up-to-date reference sequence in the genomics community. This mapping will allow clinicians to link all phenotype data on their end to the genomic data in the genomic community.
Political pros and cons have to be carefully handled and continuously explained. Ensembl openness, existing infrastructure and visibility is the biggest selling point to have these dbs linked in a common LSDB resource.
Website here will have LRG XML files and prettified HTML reports soon.
NHGRI GWAS will be imported in Ensembl: it's got manually curated data of high quality.
Links in Variation view: link "Phenotype Data (n)"
http://www.ensembl.org/Homo_sapiens/Variation/Phenotype?source=dbSNP;v=rs420259
Diagnostic testing: situation will now improve in Ensembl
Locus specific databases (LSDBs): p53, ABO, collagen, albinism, cystic
fibrosis, Altzheimer's disease, ... >700
The main aim is to be able to link the reference CDS sequence used by the biomedical community to the most up-to-date reference sequence in the genomics community. This mapping will allow clinicians to link all phenotype data on their end to the genomic data in the genomic community.
Political pros and cons have to be carefully handled and continuously explained. Ensembl openness, existing infrastructure and visibility is the biggest selling point to have these dbs linked in a common LSDB resource.
Website here will have LRG XML files and prettified HTML reports soon.
Labels: ensembl, genomics, scientific talk
Ensembl Quality Checking -- Michael Schuster
All Ensembl gene predictions for all vertebrate species are based on experimental evidence:
Aligning the evidences back to the gene prediction with Exonerate. Types of alignment results:
Exonerate has an exhaustive mode that takes a lot more time but fixes some of the mini-intron and
mini-exon issues that sometimes occur. Exonerate cdna2genome is very useful for quality checking.
Genebuild now uses head-to-head alignments of genewise and exonerate, and takes the best in each case.
Some cases are still difficult to get right with algorithmic solutions: this is were the curators are needed.
- UniprotKB/Swiss-Prot
- NCBI RefSeq proteins and mRNAs
- UniprotKB/TrEMBL
- EMBL Nucleotide Sequence Archive
Aligning the evidences back to the gene prediction with Exonerate. Types of alignment results:
- perfect
- added start
- longer region
- missing start
- non-matching start
- non-matching region
- shorter region
- ...
Exonerate has an exhaustive mode that takes a lot more time but fixes some of the mini-intron and
mini-exon issues that sometimes occur. Exonerate cdna2genome is very useful for quality checking.
Genebuild now uses head-to-head alignments of genewise and exonerate, and takes the best in each case.
Some cases are still difficult to get right with algorithmic solutions: this is were the curators are needed.
Labels: ensembl, gene prediction, genomics, scientific talk
4/07/2009
EnsemblCompara "Back to the future": using phylogenetic information to help gene annotation
Here is an example success story of using phylogenetic information to improve human gene annotation. What do you see wrong in this EnsemblCompara GeneTree?
The human gene prediction has been split into one third to the left and two thirds to the right. Some of the other species have the full length prediction, but some of the 2x and projected genomes also have this issue. This case was reported to the Havana team at the Sanger and they have now built a human and mouse full-length prediction for the gene (notice the Havana_genes blue and dark green model):
The next Ensembl/Havana merge will hopefully reflect this change but, right now, you need to activate the Havana_genes DAS track to see the most up-to-date Havana annotation. There is a good number of these fixed now in the highly loved genomes, aka human, mouse and zebrafish. But there is a second level of genomes that are not getting any manual annotation here but may be annotated somewhere else...
4/06/2009
Genomes to come: wallaby
Tammar Wallaby - Wikipedia, the free encyclopedia
The Tammar Wallaby (Macropus eugenii), also known as the Dama Wallaby or Darma Wallaby, is a small member of the kangaroo family and is the type species for research on kangaroos and marsupials.
It is found on offshore islands on the South Australian and Western Australian coast. It is classified as vermin on Kangaroo Island, where it seasonally breeds in large numbers and damages the echidna habitat on the island.
4/02/2009
Quarterly Activity Report -- Hinxton Sequence Forum Wellcome Trust Genome Campus
A quarterly activity report for the different activities that take place under what can be considered "sequence" at the WTGC in Hinxton, Cambridge, UK:
- HGNC has made great progress in solving nomenclatures for 130 cases where the community has a diversity of opinions and it's difficult to agree on something. Very good point in saying that in the Internet era it's better to give a gene a name that is distinctive to common words that would clobber your Google search results. There is now a forum set up for different communities to use in discussions for gene family names.
- Havana now has started using RNAseq data to confirm new genes found in human and zebrafish that didn't have evidence before. One new feature is a "confirmed intron" for when paired Solexa reads bridge two exons, with an associated score for read depth. Confident this type of data will bring out many interesting new annotations that couldn't be found before, e.g. genes expressed in a given tissue during a lapse of a few hours in the development. There are already a few examples in zebrafish.
- Wormbase has been working hard on compiling more data from the modENCODE. Small but cool infrastructure achievement in having VMware images running for old releases that investigators can just pull and bring up on demand.
- Ensembl Genomes has been successfully testing the beta sites for Bacteria, Protists and the first Metazoa build. Another Metazoa build is in progress, with all the phylogenetics goodness of the 12 Drosophila genomes plus the vectors plus C.elegans and a few other outgroups. The modENCODE project is about to complete the re-annotation of gene models using CAGE data that will bring more precise gene starts for melano and elegans. Ensembl Genomes is still working together with Manchester and now the US to put together an Aspergillus resource that provides the best value for money to researchers. PombBase is also being pursued, lots of labs interested in having it Ensemblified and ready to use.
3/23/2009
Laura Clarke - 1000 Genomes project
Depositing every week same amount of data as all was in public dbs before (past ~20-30 years).
Data coordination (EBI Paul Flicek, Laura+Zam). Keep submissions, run QCs and recalibration, present in mini Ensembl browser. Working on the Resembl (Solexa) public release, need to implement as MySQL 5.1 partitioning instead of commercial db.
Pipeline -- first align to the genome; will move to new assembly soon. 454 with ssaha, Solexa here with MAQ.
Trios data now being churned at dbSNP -- causing dbSNP more churning than usual releases, they are catching up. Low coverage also submitted, but will probably be in dbSNP 131.
Data formats: Fastq / BAM (binary SAM alignment map format) / GLF (genotype likelihood format). BAMs/GLFs will be updated as more data gets in and old ones will disappear.
Hope all the sequencing will be done by the end of 2009. Paper about pilot projects soon. Targetted sequencing (pilot 3) took more time, pull-down methods a bit longer to nail down, now working.
DCC more automated data delivery systems. Standard QC/Recalibration pipeline. Other high throughput analyses. New staff. May take over the alignment process once the alignment algorithm is consensuated.
Jim Stalker and Thomas Keane doing a lot of work at the Sanger. Eugene Kulesha and Stephen Keenan on the website work. Fiona and Yuan on calling/storing/presenting SNPs in Ensembl.
Data coordination (EBI Paul Flicek, Laura+Zam). Keep submissions, run QCs and recalibration, present in mini Ensembl browser. Working on the Resembl (Solexa) public release, need to implement as MySQL 5.1 partitioning instead of commercial db.
Pipeline -- first align to the genome; will move to new assembly soon. 454 with ssaha, Solexa here with MAQ.
Trios data now being churned at dbSNP -- causing dbSNP more churning than usual releases, they are catching up. Low coverage also submitted, but will probably be in dbSNP 131.
Data formats: Fastq / BAM (binary SAM alignment map format) / GLF (genotype likelihood format). BAMs/GLFs will be updated as more data gets in and old ones will disappear.
Hope all the sequencing will be done by the end of 2009. Paper about pilot projects soon. Targetted sequencing (pilot 3) took more time, pull-down methods a bit longer to nail down, now working.
DCC more automated data delivery systems. Standard QC/Recalibration pipeline. Other high throughput analyses. New staff. May take over the alignment process once the alignment algorithm is consensuated.
Jim Stalker and Thomas Keane doing a lot of work at the Sanger. Eugene Kulesha and Stephen Keenan on the website work. Fiona and Yuan on calling/storing/presenting SNPs in Ensembl.
Labels: ensembl, nextgen sequencing
3/16/2009
How much should you stretch HMM classification?
There are ongoing discussions in our Campus regarding the use of de novo clustering or HMM classification to update the family models in an Orthology database from one release to the other.
One trend that I don't favour is to use the HMM models from the previous build and classify the current protein sets to it. Then re-run the alignment and tree-building steps after that.
The other trend that I favour is to re-run the new blasts/phmmers for the new proteins, re-cluster with the other hits in the updated graph, and then re-run the alignment and tree-building steps in the new set of family models.
People who argue in favour of the HMM classification procedure and want to convince me of their feasibility show give convincing answers to these questions:
Let's say you only have a few complete genomes sequences with provisional gene predictions from your clade but expect to have 20% more extra finished genomes with better gene predictions every two months. Over the course of a year your will have more than doubled the number of genomes. Do you trust the HMMs you are doing today to represent the family models in two month, four-month, six-month, eight-month, ten-month, twelve-month time?
Let's say that you have answered the previous question with a 'yes' then, where do you draw the line to update your HMM by rebuilding them from updated genome sets? Why don't you take Human, S.cerevisiae, A.thaliana, E.coli and P.furiosus and call that the ultimate representation of all family models on Earth? Do you think those families would be as good as than the ones obtained using 80 genomes instead?
If so, then it's very easy for you, just use those 5 genomes as your family model set. But I don't think that is the way to go.
One trend that I don't favour is to use the HMM models from the previous build and classify the current protein sets to it. Then re-run the alignment and tree-building steps after that.
The other trend that I favour is to re-run the new blasts/phmmers for the new proteins, re-cluster with the other hits in the updated graph, and then re-run the alignment and tree-building steps in the new set of family models.
People who argue in favour of the HMM classification procedure and want to convince me of their feasibility show give convincing answers to these questions:
Let's say you only have a few complete genomes sequences with provisional gene predictions from your clade but expect to have 20% more extra finished genomes with better gene predictions every two months. Over the course of a year your will have more than doubled the number of genomes. Do you trust the HMMs you are doing today to represent the family models in two month, four-month, six-month, eight-month, ten-month, twelve-month time?
Let's say that you have answered the previous question with a 'yes' then, where do you draw the line to update your HMM by rebuilding them from updated genome sets? Why don't you take Human, S.cerevisiae, A.thaliana, E.coli and P.furiosus and call that the ultimate representation of all family models on Earth? Do you think those families would be as good as than the ones obtained using 80 genomes instead?
If so, then it's very easy for you, just use those 5 genomes as your family model set. But I don't think that is the way to go.
3/10/2009
What does this look like?
It's a protein multiple alignment of 871 G-protein coupled receptors visualized using Jalview Overview window. Notice the red rectangle that correspond to the bit you can actually see in detail on a 2000x1200 TFT screen, and the mouse pointer on the bottom right corner as a reference of the size of the thing.
This is one of the biggest GeneTrees in Ensembl v53, and also one of the interesting ones to navigate around.
http://www.ensembl.org/Homo_sapiens/Gene/Compara_Tree?db=core;g=ENSG00000112414
Labels: ensembl
1/12/2009
Nature vs nurture -- it's DNA methylation
Nutritional Control of Reproductive Status in Honeybees via DNA Methylation -- Kucharski et al. 319 (5871): 1827 -- Science
DNMT3 has a role in DNA methylation and genomic imprinting, with DNMT3L (EnsemblCompara GeneTree) possibly required for germ line DNA methylation. DNMT3B has been associated to lung cancer and other genetic disorders.
Fertile queens and sterile workers are alternative forms of the adult female honeybee that develop from genetically identical larvae following differential feeding with royal jelly. We show that silencing the expression of DNA methyltransferase Dnmt3, a key driver of epigenetic global reprogramming, in newly hatched larvae led to a royal jelly–like effect on the larval developmental trajectory; the majority of Dnmt3 small interfering RNA–treated individuals emerged as queens with fully developed ovaries. Our results suggest that DNA methylation in Apis is used for storing epigenetic information, that the use of that information can be differentially altered by nutritional input, and that the flexibility of epigenetic modifications underpins, profound shifts in developmental fates, with massive implications for reproductive and behavioral status.
DNMT3 has a role in DNA methylation and genomic imprinting, with DNMT3L (EnsemblCompara GeneTree) possibly required for germ line DNA methylation. DNMT3B has been associated to lung cancer and other genetic disorders.
Labels: ensembl, hard science, soft science
5/30/2008
Ensembl GeneTree for CTNNBL1: a novel gene for obesity
CTNNBL1 (Beta-catenin-like protein 1 (Nuclear-associated protein) (NAP) (Testis development protein NYD-SP19)) is a gene recently linked to obesity (HMG).
The domain structure shows an "Armadillo" SMART domain and a shorter PFAM domain
at the beginning of the protein:
The Ensembl Compara GeneTree for CTNNBL1 shows:
A high degree of conservation across a wide range of species, down to the nematode C. elegans. We unfortunately
have partial predictions for platypus, and a suspiciously partial prediction for cow, as well as the cat and shrew 2x genomes.
The domain structure shows an "Armadillo" SMART domain and a shorter PFAM domain
at the beginning of the protein:
The Ensembl Compara GeneTree for CTNNBL1 shows:
A high degree of conservation across a wide range of species, down to the nematode C. elegans. We unfortunately
have partial predictions for platypus, and a suspiciously partial prediction for cow, as well as the cat and shrew 2x genomes.
Labels: ensembl
5/19/2008
Distributed Annotation System (DAS) -- Andy Jenkinson
Two interesting points from Andy Jenkinson's DAS talk:
(a) Current clients (e.g., Ensembl) can identify that the data from an external server (Lab X) is based on an old coordinate system/assembly (OldCS), project from the old coordinate system to the new one, and prepend "OldCS_" to the names for that track.
(b) When sending BED/WIG/GFF files to current clients (e.g., Ensembl), the files are parsed, a temporary DAS is created in a server, and the data is served to the client as if it were from a DAS server. This is specially attractive for very dense feature sets that need to be visualized in queries of small segments of a chromosome. If the uploading of the file is done in compressed format, the whole operation is very efficient. This is still not the case for URL-based files, which can make your browser painfully slow.
(a) Current clients (e.g., Ensembl) can identify that the data from an external server (Lab X) is based on an old coordinate system/assembly (OldCS), project from the old coordinate system to the new one, and prepend "OldCS_" to the names for that track.
(b) When sending BED/WIG/GFF files to current clients (e.g., Ensembl), the files are parsed, a temporary DAS is created in a server, and the data is served to the client as if it were from a DAS server. This is specially attractive for very dense feature sets that need to be visualized in queries of small segments of a chromosome. If the uploading of the file is done in compressed format, the whole operation is very efficient. This is still not the case for URL-based files, which can make your browser painfully slow.
Labels: ensembl
5/15/2008
Install the Ensembl Compara API in 12 lines
cd
mkdir src
cd src
cvs -d :pserver:cvs@code.open-bio.org:/home/repository/bioperl login # password is cvs
cvs -z3 -d :pserver:cvs@code.open-bio.org:/home/repository/bioperl checkout -r bioperl-release-1-2-3 bioperl-live
cvs -d :pserver:cvsuser@cvs.sanger.ac.uk:/cvsroot/ensembl login # password is CVSUSER
cvs -z3 -d :pserver:cvsuser@cvs.sanger.ac.uk:/cvsroot/ensembl checkout -r branch-ensembl-49 ensembl
cvs -z3 -d :pserver:cvsuser@cvs.sanger.ac.uk:/cvsroot/ensembl checkout -r branch-ensembl-49 ensembl-compara
PERL5LIB=${PERL5LIB}:${HOME}/src/bioperl-live
PERL5LIB=${PERL5LIB}:${HOME}/src/ensembl/modules
PERL5LIB=${PERL5LIB}:${HOME}/src/ensembl-compara/modules
export PERL5LIB
It's easy :-)
mkdir src
cd src
cvs -d :pserver:cvs@code.open-bio.org:/home/repository/bioperl login # password is cvs
cvs -z3 -d :pserver:cvs@code.open-bio.org:/home/repository/bioperl checkout -r bioperl-release-1-2-3 bioperl-live
cvs -d :pserver:cvsuser@cvs.sanger.ac.uk:/cvsroot/ensembl login # password is CVSUSER
cvs -z3 -d :pserver:cvsuser@cvs.sanger.ac.uk:/cvsroot/ensembl checkout -r branch-ensembl-49 ensembl
cvs -z3 -d :pserver:cvsuser@cvs.sanger.ac.uk:/cvsroot/ensembl checkout -r branch-ensembl-49 ensembl-compara
PERL5LIB=${PERL5LIB}:${HOME}/src/bioperl-live
PERL5LIB=${PERL5LIB}:${HOME}/src/ensembl/modules
PERL5LIB=${PERL5LIB}:${HOME}/src/ensembl-compara/modules
export PERL5LIB
It's easy :-)
Labels: ensembl
5/13/2008
Ensembl GeneTree for NRXN1/NRXN3
Significant association of the neurexin-1 gene (NRXN1) with nicotine dependence in European- and African-American smokers
http://hmg.oxfordjournals.org/cgi/content/short/17/11/1569?rss=1
The Ensembl Compara GeneTree for NRXN1 shows:
http://www.ensembl.org/Homo_sapiens/genetreeview?db=core;gene=NRXN1
that the NRXN1 subfamily is closely related to the NRXN3 subfamily, and there is a third member, NRXN2, that is also related to the other two. The TreeFam tree shows NRXN2 on top of the tree:
http://www.treefam.org/cgi-bin/TFinfo.pl?ac=TF321302
http://hmg.oxfordjournals.org/cgi/content/short/17/11/1569?rss=1
The Ensembl Compara GeneTree for NRXN1 shows:
http://www.ensembl.org/Homo_sapiens/genetreeview?db=core;gene=NRXN1
that the NRXN1 subfamily is closely related to the NRXN3 subfamily, and there is a third member, NRXN2, that is also related to the other two. The TreeFam tree shows NRXN2 on top of the tree:
http://www.treefam.org/cgi-bin/TFinfo.pl?ac=TF321302
Labels: ensembl
3/05/2008
Ensembl GeneTree PARP (Poly (ADP-ribose) polymerase)
Loss of BRCA2 function impairs DNA repair by homologous recombination and renders cells particularly sensitive to cisplastin and also to PARP (poly (ADP-ribose) polymerase) inhibitors.
This is the Ensembl GeneTree for the PARP family, comprising PARP1, PARP2 and PARP3, showing the evolutionary history of these genes in the different species present in the Ensembl database:
http://www.ensembl.org/Homo_sapiens/genetreeview?gene=PARP1
The genetree shows one C.elegans outgroup, Y71F9AL.18, and places the insect outgroups in the PARP1 subtree, although they should probably be outgroups of all PARP genes. Treefam has the families split into PARP1, PARP2, PARP3 and so on. In the Treefam full trees, only PARP1 contains the 12 Drosophilas outgroup, in agreement with our tree, suggesting that the duplication was after the split with the insects, and PARP1 is still closer to the ancestral gene. The gene predictions for the two Cionas are not perfect, and one copy goes to PARP2 where should probably go to PARP1. There is only a Ciona savigny and no Ciona intestinalis for PARP1. There is one Tetraodon nigroviridis gene that is probably mispredicted: GSTENG00030378001, and hopefully will get better with the upcoming new gene build.
We can't see a fish-WGD effect on this tree, so the redundant copies where potentially lost after the duplication.
Technorati Tags: ensembl
This is the Ensembl GeneTree for the PARP family, comprising PARP1, PARP2 and PARP3, showing the evolutionary history of these genes in the different species present in the Ensembl database:
http://www.ensembl.org/Homo_sapiens/genetreeview?gene=PARP1
The genetree shows one C.elegans outgroup, Y71F9AL.18, and places the insect outgroups in the PARP1 subtree, although they should probably be outgroups of all PARP genes. Treefam has the families split into PARP1, PARP2, PARP3 and so on. In the Treefam full trees, only PARP1 contains the 12 Drosophilas outgroup, in agreement with our tree, suggesting that the duplication was after the split with the insects, and PARP1 is still closer to the ancestral gene. The gene predictions for the two Cionas are not perfect, and one copy goes to PARP2 where should probably go to PARP1. There is only a Ciona savigny and no Ciona intestinalis for PARP1. There is one Tetraodon nigroviridis gene that is probably mispredicted: GSTENG00030378001, and hopefully will get better with the upcoming new gene build.
We can't see a fish-WGD effect on this tree, so the redundant copies where potentially lost after the duplication.
Technorati Tags: ensembl
Labels: ensembl
Archives
200409 200412 200501 200502 200503 200504 200505 200506 200507 200508 200509 200510 200511 200512 200601 200602 200603 200604 200605 200606 200607 200608 200609 200610 200611 200612 200701 200702 200703 200704 200705 200707 200708 200709 200710 200711 200712 200801 200802 200803 200804 200805 200806 200807 200808 200809 200810 200811 200812 200901 200902 200903 200904 200905 200906 200907 200908 200909 200912 201001 201002 201003 201004 201007 201009 201011 201102
Subscribe to Posts [Atom]