EST clustering

Exercise 1: Cleaning sequences

A little clustering

Without any cleaning, we can suppose that the sequences are in the same cluster:

• seq1 overlap seq2 and seq4
• seq2 overlap seq3 and seq4

Vector clipping

Seq2 is contaminated by Phage M13 vector at his 3' end. Remove contamination sequence:

>seq2TATAAATACAAATACGTATACATGTCTATTATAATGAAAAATTGCCAATCTTGTTTAAGCAAATGCATTCTATCGTTATTATAAATGTTAGTTCTAGCTTTATTTACTTCAAAATCTTAAATCAGAATAAATTAATATTGTATTGCTGCTGTGCGTGGAAAAAGATGATGTTTATGTTCTTATAGAATAAAAGCTGTGGTTNTTTATTGTCTGTCTCCTCCACTAGANTGTAAGCTCCATGAGGGCAGGGATTTTGTCTGTYTTGTTCACTGCTGTATCCCCAGCGCCTAGAACAGTGCCTGGCACATAGTAGGCGCTCAATAAATATTTGTTGAATGAATGAATGAATGAATACGCTACTACTATTAGTAGAATTGATGCCACCTTTTCA

Seq3 is contaminated by Phage M13 vector at his 3' end. Remove contamination sequence:

>seq3TGGCACTCACCCATGGCAGCACAGGATGCCATCTTCTTTGAGTGCTGTCGTAATGAGCTGGATTNTTTATTGTCTGTCTCCTCCACTAGANTGTAAGCTCCATGAGGGCAGGGATTTTGTCTGTYTTGTTCACTGCTGTATCCCCAGCGCCTAGAACAGTGCCTGGCACATAGTAGGCGCTCAATAAATATTTGTTGAATGAATGAATGAATAATAGGACTCCTACGCTACTACTATTAGTAGAATTGATGCCACCTTTTCA

Repeat masking

Seq2 contains a LINE/L2 repeat:

>seq2TATAAATACAAATACGTATACATGTCTATTATAATGAAAAATTGCCAATCTTGTTTAAGCAAATGCATTCTATCGTTATTATAAATGTTAGTTCTAGCTTTATTTACTTCAAAATCTTAAATCAGAATAAATTAATATTGTATTGCTGCTGTGCGTGGAAAAAGATGATGTTTATGTTCTTATAGAATAAAAGCTGNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNGAAT

Seq3 contains a LINE/L2 repeat:

>seq3TGGCACTCACCCATGGCAGCACAGGATGCCATCTTCTTTGAGTGCTGTCGTAATGAGCTGGANNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNGAATAATAGGACTCCT

Seq4 contains a low complexity region:

>seq4NNNNNNNNNNNNNNNNNNNNNCATATCCCTACATCCCCTCCCTCCCACCACCCTATTCAATTGAGAGCAGGGAGATATTTTTTGGATTTCCTTTATACTGTGAAGTCACATGCATCGAAGGGTCAAACCTCTAGGTGCAGAAAAGGAAAAAAAAACCTATAAATACAAATCTGTATAGATGTCTATTATTATGAAAAATTCCCCATCTTGTTTAAGCAAATGCATTCTATCGTTATTATAAATGTTAGTTCTAGCTTTATTTACTTCCAAATCTTAAATCAGAATAAATTAATATTGTATTGCTGCTGTGCGTGGAAAAAGATGATGTTTATGTTCTTATAGAATAAAAGCTGTGGAACG

A clean little clustering

• seq1 overlap seq2 and seq4
• seq2 overlap seq4
• seq3 don't overlap with the rest of the sequences

Exercise 2: Gene Indices

Unigene

Statistics

• To get these values follow the Statistics link in the Unigene Human web page.
• mRNAs are present in the clusters because Unigene use a supervised clustering method (see course).

Keyword search

• The keyword 'myelin' returns 17 clusters.
• Cluster Hs.69547 belongs to chromosome 18 and its build using 14 mRNA sequences and 1181 ESTs.

Libraries

• The brain library 'Lib.86' contains 23 sequences. These sequences are found in 19 clusters
.

DDD

• Choose libraries 516, 517, and 768 which are labeled as 'tumor' and 'carcinoma' and compare against brest_normal.
• 2 genes are over-expressed in disease tissues: kallikrein 3 and ubiquitin C.
• IMPORTANT: Remember to check if libraries are normalized!!

TIGR Gene Indices

Statistics

• Statistics about the human database are available following the link 'Information About the Current Release'.
• There are 3,803,907 THC, the majority coming from ESTs.
• As for Unigene, TIGR Gene Indices use a supervised clustering procedure (see course).