Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jun 29;12(6):e0179529.
doi: 10.1371/journal.pone.0179529. eCollection 2017.

MDD-Palm: Identification of protein S-palmitoylation sites with substrate motifs based on maximal dependence decomposition

Shun-Long Weng  1   2   3 , Hui-Ju Kao  4 , Chien-Hsun Huang  4   5 , Tzong-Yi Lee  4   6
Affiliations

MDD-Palm: Identification of protein S-palmitoylation sites with substrate motifs based on maximal dependence decomposition

Shun-Long Weng et al. PLoS One. .

Abstract

S-palmitoylation, the covalent attachment of 16-carbon palmitic acids to a cysteine residue via a thioester linkage, is an important reversible lipid modification that plays a regulatory role in a variety of physiological and biological processes. As the number of experimentally identified S-palmitoylated peptides increases, it is imperative to investigate substrate motifs to facilitate the study of protein S-palmitoylation. Based on 710 non-homologous S-palmitoylation sites obtained from published databases and the literature, we carried out a bioinformatics investigation of S-palmitoylation sites based on amino acid composition. Two Sample Logo indicates that positively charged and polar amino acids surrounding S-palmitoylated sites may be associated with the substrate site specificity of protein S-palmitoylation. Additionally, maximal dependence decomposition (MDD) was applied to explore the motif signatures of S-palmitoylation sites by categorizing a large-scale dataset into subgroups with statistically significant conservation of amino acids. Single features such as amino acid composition (AAC), amino acid pair composition (AAPC), position specific scoring matrix (PSSM), position weight matrix (PWM), amino acid substitution matrix (BLOSUM62), and accessible surface area (ASA) were considered, along with the effectiveness of incorporating MDD-identified substrate motifs into a two-layered prediction model. Evaluation by five-fold cross-validation showed that a hybrid of AAC and PSSM performs best at discriminating between S-palmitoylation and non-S-palmitoylation sites, according to the support vector machine (SVM). The two-layered SVM model integrating MDD-identified substrate motifs performed well, with a sensitivity of 0.79, specificity of 0.80, accuracy of 0.80, and Matthews Correlation Coefficient (MCC) value of 0.45. Using an independent testing dataset (613 S-palmitoylated and 5412 non-S-palmitoylated sites) obtained from the literature, we demonstrated that the two-layered SVM model could outperform other prediction tools, yielding a balanced sensitivity and specificity of 0.690 and 0.694, respectively. This two-layered SVM model has been implemented as a web-based system (MDD-Palm), which is now freely available at http://csb.cse.yzu.edu.tw/MDDPalm/.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1
Fig 1. The conceptual diagram of constructing two-layered SVMs based on MDDLogo-identified substrate motifs.
Fig 2
Fig 2. Amino acids composition of the S-palmitoylation sites.
(A) Comparison of amino acids composition between positive data (710 S-palmitoylation sites) and negative data (5,676 non-S-palmitoylation sites). (B) Position-specific amino acids composition surrounding the S-palmitoylation sites based on frequency plot of WebLogo. (C) The compositional biases of amino acids around S-palmitoylation sites (upper panel) compared to the non-S-palmitoylation sites (lower panel) based on TwoSampleLogo (p-value < 0.01).
Fig 3
Fig 3. ROC curves of the single SVM models trained using various features based on five-fold cross-validation.
Fig 4
Fig 4. Tree-like view of MDDLogo-identified motif signatures on 710 non-homologous S-palmitoylated sequences.
Fig 5
Fig 5. ROC curves of the SVM models trained from MDDLogo-identified motifs based on five-fold cross-validation.
Fig 6
Fig 6. A case study of S-palmitoylation site prediction on human CD9 antigen (CD9_HUMAN).
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Dietrich LE, Ungermann C (2004) On the mechanism of protein palmitoylation. EMBO Rep 5: 1053–1057. doi: 10.1038/sj.embor.7400277 - DOI - PMC - PubMed
    1. el-Husseini Ael D, Bredt DS (2002) Protein palmitoylation: a regulator of neuronal development and function. Nat Rev Neurosci 3: 791–802. doi: 10.1038/nrn940 - DOI - PubMed
    1. Linder ME, Deschenes RJ (2003) New insights into the mechanisms of protein palmitoylation. Biochemistry 42: 4311–4320. doi: 10.1021/bi034159a - DOI - PubMed
    1. Smotrys JE, Linder ME (2004) Palmitoylation of intracellular signaling proteins: regulation and function. Annu Rev Biochem 73: 559–587. doi: 10.1146/annurev.biochem.73.011303.073954 - DOI - PubMed
    1. Huang K, El-Husseini A (2005) Modulation of neuronal protein trafficking and function by palmitoylation. Curr Opin Neurobiol 15: 527–535. doi: 10.1016/j.conb.2005.08.001 - DOI - PubMed

MeSH terms

Grants and funding

This work was supported by the Ministry of Science and Technology (MOST) of Taiwan under the contract numbers of 103-2221-E-155-020-MY3 and MOST 104-2221-E-155-036-MY2 to TYL. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.