\documentclass[12pt]{article}
\usepackage{amsmath,amssymb,amsfonts}
\begin{document}
We present a first global QCD analysis of next-to-next-leading-order (NNLO) contributions of the spin-dependent parton distribution functions (PPDFs) and their uncertainties using the Jacobi polynomial approach. Having the NNLO contributions of the quark-quark and gluon-quark splitting functions in perturbative QCD (Nucl. Phys. B 889 (2014) 351-400), one can obtain the evolution of longitudinally polarized parton densities of hadrons up to NNLO accuracy of QCD. A very large sets of recent and up-to-date experimental data of spin structure functions of the proton $g_1^p$, neutron $g_1^n$, and deuteron $g_1^d$ have been used in this analysis.
The predictions for the NNLO calculations of the polarized parton distribution functions as well as the proton, neutron and deuteron polarized structure functions are compared with the corresponding results of the NLO approximation. We form a mutually consistent set of polarized PDFs due to the inclusion of the most available experimental data including the recently published data from {\tt COMPASS16} experiments (Phys. Lett. B 753 (2016) 18-28).
We have performed a careful estimation of the uncertainties using the most common and practical method, the Hessian method, for the polarized PDFs originating from
the experimental errors.
The proton, neutron and deuteron structure functions and also their first moments, $\Gamma^{\rm p, n, d}$, are in good agreement with the experimental data at small and large momentum fraction of $x$. We will discuss how our knowledge of spin-dependence structure functions may be improved at small and large value of $x$ by the recent {\tt COMPASS16} measurements at CERN, the {\tt PHENIX} and {\tt STAR} measurements at RHIC, and at the future proposed colliders such as Electron-Ion collider (EIC).
\end{document}