br Protein sub cellular localization Information about

Protein sub-cellular localization Information about the sub-cellular localization of the differentially abundant proteins was gained by using the YLoc prediction system (www.multiloc.org/YLoc) based on the YLoc+ model for fungal proteins [10]. The prediction was carried out according to the biological properties of each protein and performed into 10 different locations – also searching for multiple locations – taking into account GO terms transferred from close homologous proteins. Protein localizations for each condition comparison are shown in Table 6 [1]. The complete list of all identified proteins and respective sub-cellular localization is presented in Table 3.
Acknowledgments This work was supported by the Austrian Science Fund (FWF-Der Wissenschaftsfonds, Project no. P24206) and the BOKU Equipment GmbH as the founder of the VIBT-Extremophile Center.
Background The benchmark dataset for developing computational methods to identify the methylation sites in DNA (see, e.g., [16]) is available [17], and the information thus obtained is very useful for both basic research and drug development. But so far no existing benchmark dataset whatsoever is available for developing computational methods to identify N6-methyladenosine in RNA. The present study was initiated in an attempt to construct a benchmark dataset for the later based on the experimental observations reported by Schwartz et al. [18] recently.
Data, experimental design, materials and methods The data presented here are the benchmark dataset for training and testing iRNA-Methyl [15] (http://lin.uestc.edu.cn/server/iRNA-Methyl), a web-server predictor for identifying m6A sites in the S. cerevisiae genome. By means of the m6A-seq technique, Schwartz et al. [18] first identified 1,307 methylated Rapalink-1 (m6A) sites in the S. cerevisiae genome. They have observed that most of the m6A sites share a consensus motif GAC where its center base may be methylated [18]. To construct the corresponding negative benchmark dataset, we used the flexible sliding window approach [19,20] to search the S. cerevisiae genome, and obtained 33,280 RNA segments with exactly the same GAC consensus motif that, however, were not detected by the m6A-seq technique as methylated sites. Furthermore, it had been observed via preliminary tests that when the length of the RNA segments thus derived was 51bp, the corresponding outcomes were most promising [15]. Accordingly, the 1,307 and 33,280 RNA segments each having 51bp long were designated as positive and negative samples, respectively. Also, since the size of the negative samples thus obtained is overwhelmingly larger than that of the positive samples, to minimize the false prediction caused by such a highly skewed benchmark dataset, we randomly picked out 1,307 RNA segments from the 33,280 negative samples to form a negative subset that has the same size with the positive one. The final benchmark dataset thus obtained contains 1,307 positive samples and 1,307 negative samples. Their detailed sequences are given in Appendix A. They can also be downloaded at the web-site http://lin.uestc.edu.cn/server/iRNAMethyl/data.
Conflict of interest
Acknowledgments This work was supported by the National Nature Scientific Foundation of China (Nos. 61202256 and 61301260), and the Nature Scientific Foundation of Hebei Province (No. C2013209105).
Specifications table
Value of the data
Data The data displayed here represent the outcome of micropurification steps and visualization techniques used for purifying T. b. brucei target proteins of artemisinin, which were previously detected at 60, 40 and 39kDa by immunoblotting [1]. The polyacrylamide gel and immunoblotting film presented here below reflect, on the one hand, the successful isolation of the 60-kDa protein band but, on the other hand, the difficulty to isolate both low-abundance proteins at 40 and 39-kDa. It should be noted that a two-dimensional SDS-PAGE for further purity assessment of the 60-kDa target protein band was not yet performed.