Apollo Exercises Kansas State University 2015
0 likes276 views
The document outlines a series of exercises conducted during a Web Apollo workshop at Kansas State University, focusing on the analysis and annotation of honey bee gene models fragmented in current gene sets. Each exercise involves comparing gene models across species, revising exon structures based on RNA-seq evidence, and ensuring the integrity of protein sequences through various bioinformatics tools. Key tasks include merging fragmented genes, adjusting coordinates for canonical splice sites, and checking the accuracy of final annotations against public databases.
Apollo Exercises Kansas State University 2015
- 1. # Web Apollo Workshop - Exercises # At Kansas State University # 17 June, 2015 # M. Munoz-Torres # Berkeley Bioinformatics Open Source Projects # Lawrence Berkeley National Laboratory ----------------------------------------------------------------------- # Web Apollo Workshop - KSU # Exercise 1 It has come to our attention that the Apis mellifera (honey bee) ortholog of Nesprin-1 (nuclear envelope spectrin repeat proteins) or Muscle specific protein 300 is fragmented into possibly 5 different genes in the current gene set. In Drosophila melanogaster MSP-300 anchors nuclei to actin, and has been reported to be essential for positioning of nurse cell nuclei during oogenesis, and thus production of mature ooctyes. More information available at http://en.wikipedia.org/wiki/Nesprin You may use this fragment from the Megachile rotundata (the alfalfa leafcutter bee) model (NCBI identifier XP_003705060.1; at http://www.ncbi.nlm.nih.gov/) to pull the corresponding location in Apis mellifera using BLAT. The protein has over 12K amino acid residues in M. rotundata. >M_rotundata nesprin-1-like (fragment) MRIVEGRYSGSGGYWNVTILEGGDTGTWGYQKVGILERGGTRRWGYWNVGVPEGGDTGTWGYQKVGILER GGTRRWGYWNVGVPEGGDTGTWGYQKVGILERGGTRRWGYWNVGVPEGGDTGTWGDSGAWGYWNVRILEG GDTGTWRYQKVWSGGRTASSEELFQELDGRRNVGHFRSPTNPRPSSRASDSSFEESFERLVEEGELNGAK VVKFEKITVRKSVREVAGTGVSHQRVLAETSRTPSEEHALEDSAYQSHSHGAPSHGSKSSSVTSFTRFPS EESLSQRRGSSPQQHLGPDDRTPSEWYAEYHTQSFQNVAARIEYVRSKSEYDAHIAEIKDEQERVQKKTF VNWINSYLSKRIPPLRVDDLIDDLKDGTRLLALLEVLSGEKLPVERGRNLKRPHFLSNANTALQFLQS Gene models GB40006-RA and GB40007-RA have significant sequence similarity with this gene model. Using the entire M. rotundata model, shows high-scoring segment pairs (hsp) across GB40006- RA, GB40007-RA, GB40008-RA, GB40009-RA, and GB40010-RA. Clear highlight. Drag and merge GB40006-RA, GB40007-RA, and GB40008-RA Rename transcript to: nesprin-1-like-RA Observing Nurse RNA-seq reads, there are no reads in support of the exon at 483,800 - 483,705, while there are many reads in support of an intron passing across the region. Delete the exon (#1 of GB40007-RA). Closer inspection of RNA-seq reads also reveals that there are many reads in support of NOT having an exon in coordinates 494,235 - 491,660. Delete exon (#2 of GB40006-RA). It is now time to inspect the adjacent exons at the merging of GB40007-RA and GB40008-RA. Guided by RNA-seq reads (their length, not their exact coordinates), the phase and the structure of all exons downstream (3') of 454,041 (… 453,654), you may adjust the coordinates for the adjacent exons in this region, so they may have canonical splice sites. Monica Munoz-Torres, PhD Lead Biocurator | Bioinformatics Analyst Lawrence Berkeley National Laboratory Joint Genome Institute for the US Department of Energy 2800 Mitchell Dr. Ste 100. Walnut Creek, CA 94598 mcmunozt@lbl.gov | @monimunozto | (925) 296-5624
- 2. First, the last exon of GB40007-RA can be adjusted to the nearest canonical splice, located 5' in position 454,041. There is a gap between 453,759 .. 453,710. The first exon of GB40008-RA should also be adjusted to the nearest canonical site that extends the protein and is supported by RNA-seq reads. After trying a few sites, position 453,659 is a canonical splice site that conserves the phase of the rest of the product and matches the boundaries of RNAseq reads (-2 -- f 1,2,3). There are many exons in this model; compared to the previous bee model for A. mellifera (in NCBI), there is an excess of about 20 exons. This newly created gene product is 7,600 amino acids in length. Use this product to query NCBI's nr for orthologs, using BLAST at http://blast.ncbi.nlm.nih.gov/Blast.cgi The search retrieves homologs from other Apidae (A. dorsata, A. florea, N. vitripennis, B. impatiens) of similar length --> ~39Kbp, 12K aa. Modify the exon coordinates at the flagged non-canonical sites at the junctions of gene models using RNAseq to guide the changes and use "color by CDS" feature to preserve the phase. RNA-seq shows that the last exon in GB40008-RA may not be real (or perhaps, it is possible that there are more than one isoform (at 431,750)). There are 11 isoforms in D. melanogaster. With RNA-seq evidence as support, use Apollo’s 'edge-matching' functionality to modify the 5' end of the first exon of GB40009-RA to match the boundary of transcripts from 431,668 to that matching RNAseq evidence at 431,683 (choose exon and RNA-seq read, right-click, 'set as 5' end'). Remember: only those reads oriented in the same sense as the transcription will work with the ‘set as x’ end’ operation. Name this isoform (RA) and create a copy using the option "Duplicate" from the right-click menu. In the second isoform (RB), delete exon at ~431,750 (last exon of GB40008-RA) to reveal that the phase in the rest of the peptide is restored to that of original gene model. Copy the 9009 aa residues and query the public databases again using BLAST at NCBI. Merge the model with GB40010-RA, rename to reflect that this is isoform B. At the junction, follow RNA-seq reads pattern to adjust the last exon of GB40009-RA to canonical splice site at 427,372, and the first exon of GB40010-RA to canonical splice site at 427,069 immediately 3' of the GAP and the longest extension supported by RNA-seq data. You may also use additional evidence tracks to support these decisions: e.g. Forager Illumina Bee Contigs, Mixed Antennae 454 Contigs, and Fgenesh++ with RNAseq training data. In this peptide, at position 394,197 a GC splice donor (common in insects) is visible and heavily supported by the evidence. (GC/AG instead of GT/AG). Note on the comments and finish the annotation. A final product of 16,509aa in length is obtained. Check integrity and accuracy: - blastp vs Insecta (at NCBI) (-Apis mellifera ; i.e. without honey bee) - add GO terms (from Drosophila entries - PAINT) - search the literature to add PubMed IDs (http://www.ncbi.nlm.nih.gov/pubmed).
- 3. ----------------------------------------------------------------------- # Web Apollo Workshop - KSU # Exercise 2 It has come to our attention that the honey bee ortholog of small ribonucleoprotein particle protein SmD2 is fragmented into 2 or more genes in the current gene set. In Drosophila melanogaster SmD2 are involved in mRNA splicing, via spliceosome. You may use the Apis florea model (XP_003697276.1) to pull the corresponding location in Apis mellifera using BLAT. >A_florea Sm D2-like MLNNIRNITYFSDQSSLTKPKSEMTPEELAKREEEEFNTGPLSVLTQSVKNNTQVLINCRNNKKLLGRVK AFDRHCNMVLENVKEMWTELPRTGKGKKKAKPVNKDRFISKMFLRGDSVILVLRNPLATASGK Drag the honey bee gene model you found into the 'User-created Annotation Area' and retrieve its sequence to query the public databases with it to inspect their structure. Query NCBI's nr for orthologs, using BLAST at http://blast.ncbi.nlm.nih.gov/Blast.cgi Review conserved domains, compare the alignments with other sequences and inspect their length differences. Find evidence in support of this model looking through the transcriptomes of Forager bees and Nurse bees as well as available ESTs. For now, please refrain from using previous bee gene models. RNA-seq evidence will support current exons, as well as an additional exon in the 5'end. Check integrity and accuracy: - blastp vs Insecta (at NCBI) (-Apis mellifera) - add GO terms (e.g. from Drosophila entries) - search the literature to add PubMed IDs (http://www.ncbi.nlm.nih.gov/pubmed). After checking the current protein product vs public databases, alignments will show a discrepancy in the 5'end. Can you spot the difference? Are you able to find the missing residues in the current A. mellifera genome assembly? Is there enough evidence to support this hypothesis? How many isoforms would you annotate in this case?
- 4. ----------------------------------------------------------------------- # Web Apollo Workshop - KSU # Exercise 3 It has come to our attention that the honey bee ortholog of FlyBase model CG31619 is fragmented into 2 or more genes in the current gene set. In Drosophila melanogaster CG31619 is involved in proteolysis, and has metalloendopeptidase activity (molecular function). We know that the homolog of CG31619 in Apis dorsata is similar to the ADAMTS proteins. ADAMTS = A Disintegrin And Metalloproteinase with Thrombospondin Motifs, which is a family of peptidases. You may use the Apis dorsata model (XP_003697278.1) to pull the corresponding location in Apis mellifera using BLAT in Apollo. >Apis_dorsata ADAMTS-like protein partial RENPRYWLQPCENPSDFRSEQCAAFDDVPYSGQLLKWYPHYDPSRPCALICRGEQSVENTGNRLRQE TSVEKTLPHDATDALQLDSEETIVVQLADKVEDGTKCYTDSMDVCINGECMKVGCDLRVGSNKNTDPCGV CGGNGSSCQSRYSWSLESISACSKSCGGGFKIAMAVCKAIGPDESVVDDSYCDPDNRPEKTLMPCNTHPC Drag each of the honey bee gene models you found into the 'User-created Annotation Area' and retrieve their sequences to query the public databases with them to inspect their structure. Query NCBI's nr database for orthologs, using BLAST at http://blast.ncbi.nlm.nih.gov/Blast.cgi Review conserved domains, compare the alignments with other sequences and inspect their length differences. Once you inspect the expected conserved domains, it will be apparent that the models need to be brought together. Find evidence in support of this operation looking through the transcriptomes of Forager bees and Nurse bees as well as available ESTs. For now, please refrain from using previous bee gene models. The available evidence will show support of splice site corrections and possible excess of exons. Once you are satisfied with a final model, check integrity and accuracy: - blastp vs Insecta (at NCBI) (-Apis mellifera) - add GO terms (e.g from Drosophila entries) - search the literature to add PubMed IDs (http://www.ncbi.nlm.nih.gov/pubmed).
- 5. ----------------------------------------------------------------------- # Web Apollo Workshop – KSU # Exercise 4 It has come to our attention that the honey bee ortholog of RNA Polymerase II 140 KD is fragmented into 2 or more genes in the current gene set. In Drosophila melanogaster RNA Pol II is involved in transcription from RNA polymerase II promoter and has DNA binding and DNA-directed RNA polymerase activity (molecular functions). You may use the Apis dorsata model (XP_006610182.1) to pull the corresponding location in Apis mellifera using BLAT in Apollo. >Apis_dorsata RNA pol II subunit RPB2-like partial MYSLEEDQYDDEDAEEISSKLWQEACWIVINAYFDEKGLVRQQLDSFDEFIEMSVQRIVEDSPQIDLQAE AQHTSGEIENPVRHLLKFEQIYLSKPTHWEKDGAPSPMMPNEARLRNLTYSAPLYVDITKTIVKDGEDPI ETQHQKTFIGKIPIMLRSKYCLLAGLSDRDLTELNECPLDPG Drag each of the honey bee gene models you found into the 'User-created Annotation Area' and retrieve their sequences to query the public databases with them to inspect their structure. Query NCBI's nr for orthologs, using BLAST at http://blast.ncbi.nlm.nih.gov/Blast.cgi Check to notice the conserved domains -- are they intact in each model? Will bringing these models together improve the integrity of the RNA Pol II subunit RPB2 homolog in honey bee? Re-arrange the boundaries at the joining region to be canonical splice sites on each side. Did you spot the sequence error visible in the DNA-Track? What effect do you think this had on the automated annotation process? Once you are satisfied with a final model, check integrity and accuracy: - blastp vs Insecta (at NCBI) (-Apis mellifera) - add GO terms (e.g. from Drosophila entries) - search the literature to add PubMed IDs (http://www.ncbi.nlm.nih.gov/pubmed).
- 6. ----------------------------------------------------------------------- # Web Apollo Workshop - KSU # Exercise 5 It has come to our attention that the honey bee ortholog of Ceramidase is fragmented into 2 or more genes in the current gene set. Ceramidase is an enzyme, which cleaves fatty acids from ceramide, producing sphingosine (SPH) which in turn is phosphorylated by a sphingosine kinase to form sphingosine-1-phosphate (S1P). Ceramide, SPH, and S1P are bioactive lipids that mediate cell proliferation, differentiation, apoptosis, adhesion, and migration. You may use the Bombus terrestris model (XP_003397164.1) to pull the corresponding location in Apis mellifera using BLAT. The following is a fragment. >B_terrestris Ceramidase-like GTTTAAGAGTGTTCGCGCCAATTGTTCGCGGCGAGACTGGCCGTGCAGACCAGCTGTTATAGCCGCGTCT CCGCTCTGTCTCTGCTGATCCATCGATCACCTACGCATCGATCCCTCGTTCGATCAACGTGGTCATGAGC TGGAGCGTTTGAGCGCCGCTATCAGACTGGCGGCAGAGAAAAACTGAATGGAGGCACCGGCAGTTGGACG CTTTAGAATCCTTGCGTTGTTGACGATATGGCTGGTCCAGCTTGCGGTGCCCGGCGCCATCGCGTCTTAC AGCATCGGGGTGGGCAGAGCAGATGCTACAGGACCCGCCGCTGAAATTGTTTTTATGGGCTACGCGAAGA TCGATCAAAAAGGATCAGGAATCCATCTTCGAACATTCTCCCGCGCATTCATCATCGACGATGGCGAGGA GAGGTTCGTCTTCGTCAGCGTGGATAGCGCCATGATAGGAAACGGCGTTCGTCAAACGGTGTTGCAGAAT CTTGAAAAGGAGTTTGGCAGCCTGTACACAGAGAAAAATGTGATGATCAGTGCAACTCACTCGCACTCCA CACCCGGTGGATTCATGTTGCACATGTTGTTCGATATTACGACATTCGGTTTCGTTCAAGAGACCTTCGA TGCTATGGTCAAGGGAATCACGAAGAGTATTCAACGTGCTCACTATGCCATAGTTCCAGGCAGAATATTC ATCACCCATGGAGAAGTTCATGGTGTGAACATTAATAGAAGCCCATCCG Drag the honey bee gene model(s) you found into the 'User-created Annotation Area' and retrieve its sequence(s) to query the public databases to inspect its/their structure. Query NCBI's nr for orthologs, using BLAST at http://blast.ncbi.nlm.nih.gov/Blast.cgi Check to notice the conserved domains, if any. And if so, are they intact in each model? After inspecting the biological evidence in the area, do you think these models should be brought together? If so, how? Are you able to establish whether you should annotate UTRs to this gene model hypothesis? Once you are satisfied with a final model, check integrity and accuracy: - blastp vs Insecta (at NCBI) (-Apis mellifera) - add GO terms (e.g. from Drosophila entries) - search the literature to add PubMed IDs (http://www.ncbi.nlm.nih.gov/pubmed).
- 7. ----------------------------------------------------------------------- # Web Apollo Workshop - KSU # Exercise 6 It has come to our attention that the honey bee ortholog of adenomatous polyposis coli (APC) - like is fragmented into possibly 4 or more genes in the current gene set. In Drosophila melanogaster adenomatous polyposis coli (APC) -like is involved in beta-catenin binding and microtubule binding. APC is a negative regulator that controls Beta-catenin concentrations and interacts with E- cadherin, which are involved in cell adhesion. APC is classified as a tumor suppressor gene. Mutations in the APC gene may result in colorectal cancer. You may use the Camponotus floridanus gene model (EFN68235.1) to pull the corresponding location in Apis mellifera using BLAT. The following is a fragment. >Camponotus floridanus APC-like partial GRVLCGSAASVGVEGVGGAWGAQRRQSPPPLASRRLESAADAVVAAAAAAAVATTAASFNPVTQRCAKKG AVDAVAPRRIYPGPPLVAETSKEDAMEPEAKEEDERRSSTPSPEFYLRRSKRYNEDGSSEEETKQDSISL PNHRLPSSYRGTWPIRRDVWANQQIGFPAQQSTSLAAHNDVASVMSFSSSSNSAGLDSHSVELQGHRRLG AKVDVVYNLLGMLEKNGRDDMSTTLLSMSTSLDSCLVMRQSGCLPLLVQLIHAPRQDPDTRDRAMQALHN VVHAKSDERAGRREARVLRFLEQLRDYCQSLRMSLERGQSMDDLERGHPAATIAALMKLSFDEAHRHAMC QLGGLHAVAELIEMDHLAHGSESDDQNCITLRRYAGMALTNLTFGDGNNKALLCSFKEFMKALVSQLKSP SDDLRQVTASVLRNLSWRADSSSKQTLREVGAVTGLMKAAMEGRKESTLKSILSALWNLSAHCSTNKVDI Check to notice the conserved domains, if any. And if so, are they intact in each model? After inspecting the biological evidence in the area, do you think these models should be brought together? if so, how? Once you are satisfied with a final model, check integrity and accuracy: - blastp vs Insecta (at NCBI) (-Apis mellifera) - add GO terms (e.g. from Drosophila entries) - search the literature to add PubMed IDs (http://www.ncbi.nlm.nih.gov/pubmed).