Sentence examples for terms assignments from inspiring English sources

One supplementary concern about the ML classifier was: are all GO terms assignments accumulated during the past useful to profile a just published abstract?

Thanks to the annotation date contained in our three-tuples, we were able to discard all GO terms assignments inserted after a given year.

Yet, the quality of the GO terms predicted by machine learning continues to improve, thanks to the growing number of high-quality GO terms assignments available in GOA: since 2006, GOCat performances have improved by ∼50%.

Hence, we only considered manual curation linked to a PMID and discarded the gene product references in order to obtain ∼280 000 GO terms assignments expressed by three-tuples (GO id; PMID; annotation year) such as (GO 2000032; 17307920102010).

In the entire knowledge base, the 278 000 GO terms assignments concerned 20 000 distinct GO terms; yet, the number of annotation examples for each GO term was very imbalanced, with 13.8 for mean but 3 for median.

Further experiments (not reported) showed that the knowledge base must clearly contain all previous GO terms assignments; it appears that the largest the knowledge base we exploited, the most accurate were the predicted GO terms.

The number of transcripts with GO terms assignment was between 54.8 60.8% (Table 4).

Our study showed that machine learning approaches are now nearly able to reproduce the quality of GO terms assignment as performed by trained human curators.

The customized cutoff values take effect to all kinds of sequence-similarity based analysis in TrED, such as GO terms assignment and metabolic pathway reconstruction.

Then, we describe the two competing approaches for performing the automatic GO terms assignment task: the thesaurus-based classifier (EAGL) and the machine learning classifier (GOCat).

Every week, write down a list of your assigned tasks — short-term assignments and long-term goals — and rank them by importance, from your perspective.