| 1. |
- Lindlöf, Angelica
(författare)
-
Gene identification through large-scale EST sequence processing
- 2003
-
Ingår i: Applied Bioinformatics. - : Adis International. - 1175-5636. ; 2:3, s. 123-129
-
Forskningsöversikt (refereegranskat)abstract
- The technology of sequencing expressed sequence tags (ESTs) offers a relatively cheap alternative to whole genome sequencing and has become a valuable resource for gene discovery. The inherent characteristics of ESTs, such as transcript redundancy, low sequence quality and high error rates, require processing of the sequences before any gene prediction can be made. The process includes EST pre-processing, analysis and similarity searches, and the data are generally stored in a database to organise the results and thereby assist the search for interesting genes.
|
|
| 2. |
- Lindlöf, Angelica, et al.
(författare)
-
Genetic network inference : the effects of preprocessing
- 2003
-
Ingår i: Biosystems (Amsterdam. Print). - : Elsevier. - 0303-2647 .- 1872-8324. ; 72:3, s. 229-239
-
Tidskriftsartikel (refereegranskat)abstract
- Clustering of gene expression data and gene network inference from such data has been a major research topic in recent years. In clustering, pairwise measurements are performed when calculating the distance matrix upon which the clustering is based. Pairwise measurements can also be used for gene network inference, by deriving potential interactions above a certain correlation or distance threshold. Our experiments show how interaction networks derived by this simple approach exhibit low—but significant—sensitivity and specificity. We also explore the effects that normalization and prefiltering have on the results of methods for identifying interactions from expression data. Before derivation of interactions or clustering, preprocessing is often performed by applying normalization to rescale the expression profiles and prefiltering where genes that do not appear to contribute to regulation are removed. In this paper, different ways of normalizing in combination with different distance measurements are tested on both unfiltered and prefiltered data, different prefiltering criteria are considered.
|
|
