The composition of a conventional fuel can be estimated thanks to the API (Riazi-Daubert) relations linking the hydrocarbon group fractions with well-chosen properties. These relations cover only the paraffin (coupling iso and normal), naphtene and aromatic (PNA) groups as they were developed for conventional fuels presenting neglected amounts of olefin and oxygenate. However, the last quoted groups can be present in unconventional fuels. This paper presents a method to estimate the n-paraffin, iso-paraffin, olefin, naphtene, aromatic and oxygenate (PIONAOx) composition of an automotive shredder residues-derived gasoline-like fuel. Whereas the existing methods rely on the resolution of a system of three equations based on two properties among the refractivity intercept, the viscosity gravity function (VGF), the specific gravity (SG), the m-parameter, and the carbon-to-hydrogen weight ratio (HC), the current work hinge on an optimization problem relying on six properties - namely the refractive index, the VGF, the Watson K-factor, the SG, the stoichiometric air-to-fuel ratio and the oxygen content - and depending on two inputs which are the molecules selected to describe each hydrocarbon group and the weights involved in the objective function.The properties of each hydrocarbon group are calculated with the pseudo-component method rather than using an averaged value over the boiling point range. The accuracy of the method was investigated for different inputs, comparing the calculated composition with the confident intervals of the real fractions measured with a two-dimensional gas chromatography (GCxGC) technique. The lowest cumulated absolute error comparing with the confident interval of the measured fractions was evaluated at 7%. This error is due to a bad estimation of the iso-paraffins (error of 6%) while the n-paraffins, the olefins, and the oxygenates were estimated within the confident interval.