After extrusion the samples were collected, cooled to room temperature under natural convection conditions. The samples were then milled to a 0.149 mm granule size. They were labeled as extruded amaranth flours

and kept at 10 °C until analysis. Untreated flours were stored in the same manner as the extruded samples. In order to assess possible effects of flour particle size on the analysis, granule sizes were checked using a Malvern Mastersizer S-MAN 5005 (Malvern Instruments Ltda, Malvern, UK). (data not shown). The chemical composition of the flours including the moisture, fat, protein and ash content, were determined by the method described in AOAC (1997). The dietary fiber was analyzed using the enzymatic and gravimetric method according to Prosky, Asp, Schwiser, Devries, and Furnas (1988); starch was determined according to the this website method of Rickard and Behn (1987). Starch was quantified

by enzymatic hydrolysis as described by Rickard and Behn (1987). Amylose content was determined following the method ISO 6647 (International Organization for Standardization, 1987). Amylopectin content was equal to the value obtained by subtraction of amylose from total starch. The color of the samples was determined in triplicate using the equipment ColorQuest XE (Hunter Lab, ColorQuest, USA). The CIE L∗a∗b∗ system was employed. This system determines the L∗, a∗ and b∗ values, where L∗ represents lightness with 0 for black and 100 for white;

a∗ represents the opposition between green and ROCK inhibitor red colors ranging from positive (green) to negative (red) values; and b∗ is the yellow/blue opposition also ranging from positive (yellow) to negative (blues) values. In the CIE L∗a∗b∗ color space a∗ and b∗ values exhibit minima and maxima values that depend on L∗ value. To determine the water absorption Adenosine (WAI) and the water solubility indexes (WSI), the methodology proposed by Anderson, Conway, and Griffin (1969) was followed. Pasting properties of amaranth flours were determined using a Rapid Visco Analyzer (RVA-4, Newport Scientific, Warriewood, Australia) according to Ragaee and Abdel-Aal (2006). The pasting temperature (PT), peak viscosity (PV, the maximum hot paste viscosity), holding strength or trough viscosity (the trough at the minimum hot paste viscosity), final viscosity (FV, the viscosity at the end of test after cooling to 50 °C and holding at this temperature), breakdown (BD, peak viscosity − holding strength or trough viscosity) and setback (SB, final viscosity − holding strength) were determined with Thermocline for Windows software (Version 2.0). The viscosities are presented in Rapid Visco Units (RVU). Thermal properties were analyzed using a Differential Scanning Calorimeter (DSC822, Mettler Toledo, Schwerzenbach, Switzerland) according to González, Carrara, Tosi, Añón, and Pilosof (2007) with some modifications. Amaranth flour (13.0 ± 0.